United States v. Jabree Williams

PRECEDENTIAL

UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT
__________

Nos. 17-2111, 17-3191, 17-3373, 17-3586, 17-3711,
17-3777, 18-1012, 18-1324, 18-2468 and 19-1037
__________

UNITED STATES OF AMERICA

v.

JABREE WILLIAMS,
a/k/a “MINUTE”
Appellant in
No. 17-2111

ROLANDO CRUZ, JR.,
Appellant in
No. 17-3191

MARC HERNANDEZ,
a/k/a Marky D.
Appellant in
No. 17-3373

ROSCOE VILLEGA,
Appellant in
No. 17-3586
 EUGENE RICE,
also known as “B MOR”
Appellant in
No. 17-3711

DOUGLAS KELLY,
Appellant in
No. 17-3777

ANGEL SCHUEG, a/k/a “POCKO”
Appellant in
No. 18-1012

MAURICE ATKINSON,
Appellant in
No. 18-1324

ANTHONY SISTRUNK
a/k/a “KANYE”
Appellant in
No. 18-2468

TYREE EATMON, a/k/a Ree,
Appellant in
No. 19-1037
__________

On Appeal from the United States District Court
for the Middle District of Pennsylvania
(D. C. Nos. 1-14-cr-00070-017:1-14-cr-0070-004;
1-14-cr-0070-001; 1-14-cr-00070-003;1-14-cr-00070-011;1-
14-cr-0070-002;1-14-cr-00070-012; 1-14-cr-00070-008; 1-
14-cr-00070-009;1-14-cr-00070-006)

2
 District Judge: Honorable Yvette Kane

__________

Argued December 10, 2019
Before: RESTREPO, ROTH and FISHER, Circuit Judges.

(Filed: September 10, 2020)

Jonathan W. Crisp
Crisp & Associates
4031 North Front Street
Harrisburg, PA 17110
Counsel for Jabree Williams

Jeremy B. Gordon
Suite 106
1848 Lone Star Road
Mansfield, TX 76063
Counsel for Rolando Cruz, Jr.

Peter Goldberger [ARGUED]
50 Rittenhouse Place
Ardmore, PA 19003
Counsel for Marc Hernandez

Edson A. Bostic, Federal Public Defender
Tieffa N. Harper

3
Office of Federal Public Defender
800 King Street, Suite 200
Wilmington, DE 19801
Counsel for Roscoe Villega

G. Scott Gardner
2117 West 4th Street
Williamsport, PA 17701
Counsel for Eugene Rice

Richard F. Maffett, Jr
2201 North Second Street
Harrisburg, PA 17110
Counsel for Douglas Kelly

Terrence J. McGowan
Killian & Gephart
218 Pine Street
P.O. Box 886
Harrisburg, PA 17108
Counsel for Angel Schueg

John F. Yaninek [ARGUED]
Thomas Thomas & Hafer
305 North Front Street, 6th Floor
Harrisburg, PA 17101
Counsel for Maurice Atkinson

Daniel M. Myshin [ARGUED]
P.O. Box 33
Hummelstown, PA 17036
Counsel for Anthony Sistrunk

4
Andrew J. Shubin
333 South Allen Street
State College, PA 16801
Counsel for Tyree Eatmon

David Freed, United States Attorney
Michael A. Consiglio [ARGUED]
Office of United States Attorney
228 Walnut Street, P.O. Box 11754
220 Federal Building and Courthouse
Harrisburg, PA 17108
Counsel for Appellee
__________

OPINION OF THE COURT
__________

FISHER, Circuit Judge.
In mid-September 2014, a federal grand jury in the U.S.
District Court for the Middle District of Pennsylvania returned
an indictment of twenty-one men from the South Side neigh-
borhood of York, Pennsylvania. All twenty-one were charged
on counts of racketeering conspiracy, drug-trafficking conspir-
acy, and drug trafficking. Four were also variously charged
with federal firearms offenses related to the alleged trafficking.
Although so called because of its geographic location in the
city, South Side, the indictment alleged, had constituted since
2002 the identity of a criminal enterprise associated through its
upper echelons with the Bloods, a national street gang. At the
heart of the enterprise, it was said, lay an extensive drug-traf-

5
ficking operation, conducted across a defined territory and nur-
tured in part through sporadic episodes of occasionally deadly
violence involving rival gangs, gang affiliates, and, collater-
ally, members of the general public.
Over the course of the ensuing year, several of the de-
fendants pleaded guilty. Twelve, however, proceeded to a joint
trial, held over eight weeks from September to November
2015. The jury heard from well over one hundred witnesses,
including some of the original twenty-one who chose to coop-
erate with the Government in the hope of a reduced sentence.
The picture that emerged was of lives characterized by cycles
of crime and incarceration, stretching across more than a dec-
ade and punctuated by moments of significant and sometimes
reckless violence. The witnesses depicted widespread drug
dealing in crack cocaine and heroin. They told of territorial ri-
valries, market competition, and personal feuds. They re-
counted episodes of threat and retaliation, attack and retribu-
tion. But they also described friendship, loyalty, and loss; pride
and fear; ambition, and great ability left unrealized. In the end,
all twelve defendants were convicted on one or more of the
charges against them, and in the years thereafter were sen-
tenced to, among other things, terms of imprisonment ranging
from sixty months to life.
Ten of the twelve (the Defendants) now appeal their
convictions and sentences on a variety of grounds, advanced
both severally and collectively. These issues, which span more
or less all the relevant phases of a criminal prosecution, can be
divided into five categories. First, most of the Defendants con-
tend that because the District Court’s closure of the courtroom
to the public during jury selection violated their Sixth Amend-
ment right to a public trial, their convictions should be reversed
and a new trial ordered under Federal Rule of Criminal Proce-
dure 52(b). Second, two Defendants claim that the District

6
Court’s in camera disposition of a challenge under Batson v.
Kentucky, 476 U.S. 79 (1986), both violated their constitu-
tional right to personal presence at all critical phases of their
criminal trial and was sufficiently prejudicial to warrant rever-
sal of their convictions. Third, several Defendants bring evi-
dentiary challenges. Two appeal the District Court’s denial of
their motions to suppress evidence collected from their resi-
dences pursuant to search warrants. Still more Defendants as-
sert various errors regarding the admission and use of evidence
at trial. Fourth, nearly all the Defendants contend that the evi-
dence was insufficient to support one or more of the verdicts
against them. These challenges ask us to clarify, among other
things, the effect of our recent decision in United States v.
Rowe, 919 F.3d 752 (3d Cir. 2019)—and thereby of the Su-
preme Court’s decision in Alleyne v. United States, 570 U.S.
99 (2013)—upon our case law regarding the elements of a
drug-trafficking conspiracy under 21 U.S.C. § 846. Finally, all
the Defendants appeal their sentences, principally alleging pro-
cedural defects in the District Court’s judgments.
For the reasons that follow, we will affirm the Defend-
ants’ judgments of conviction. We will also affirm the judg-
ments of sentence of Jabree Williams and Eugene Rice. But we
will vacate either in whole or in part the other Defendants’
judgments of sentence, and remand the cases of Marc Hernan-
dez and Angel Schueg for resentencing proceedings consistent
with this opinion.

7
 I. BACKGROUND
A. Investigation and Indictment
These cases began with an act of cooperative federal-
1
ism. At the initiation of, and together with, local law enforce-
ment, the federal Bureau of Alcohol, Tobacco, Firearms and
Explosives (ATF) conducted a multiyear investigation into
drug trafficking and violence in the city of York, Pennsylvania.
The investigation centered on what the Government called “the
Southside Gang,” after the neighborhood in which it was said
to operate. Over the first decade of the current century, York
law enforcement officials perceived in the city a pattern of es-
calating violence that they attributed primarily to a rivalry be-
tween the South Side and Parkway, another supposed gang,
named for a public housing project in the northern part of York.
The Government associated this violence, which also occa-
sionally involved other neighborhood groups, with the wide-
spread drug trafficking throughout the South Side. It was be-
lieved that the principal sources of these drugs—and concom-
itantly of the increased violence—were individuals affiliated
with the Bloods, who had developed the South Side’s existing
drug trafficking into a more organized operation.
Legal proceedings began in mid-March 2014, when a
grand jury in the Middle District of Pennsylvania returned an
indictment of three men, Hernandez, Roscoe Villega, and
Douglas Kelly, charging them on counts of drug-trafficking
conspiracy and drug trafficking. Shortly thereafter, govern-

1
We provide here a broad overview of the cases’ factual and
procedural background, with particular attention to the five cat-
egories of issues described above. More detailed factual de-
scription will be provided where relevant in Parts II-VI below.

8
ment officials obtained and executed search warrants for sev-
eral locations across York, seizing (among other things) drugs,
drug paraphernalia, cellphones, and money. Some of this evi-
dence, as well as some seized later, became the subject of an
ongoing contest between the parties. Hernandez, Villega, and
Kelly all pleaded not guilty, but before they could proceed to
trial, a superseding indictment added Rolando Cruz, Jr. to the
list of defendants and supplemented the drug counts with two
federal firearms charges. Cruz also pleaded not guilty, but yet
again, before a trial could occur, matters developed further.
In September, the grand jury returned a second super-
seding indictment that vastly expanded the scope and ambition
of the prosecution. The indictment now listed twenty-one de-
fendants, including the original four. It charged all twenty-one
on three counts: (I) conspiracy to violate 18 U.S.C. § 1962(c)
of the Racketeer Influenced and Corrupt Organizations Act
(RICO), 18 U.S.C. § 1962(d); (II) conspiracy to distribute a
controlled substance, 21 U.S.C. § 846; and (III) distribution of
a controlled substance, 21 U.S.C. § 841(a). Counts II and III
specified drug quantities of 5 kilograms or more of powder co-
caine, and 280 grams or more of crack cocaine.2 Distribution
at these quantities carries increased penalties. See 21 U.S.C.
§ 841(b)(1)(A). The indictment also included vestiges of its
earlier iterations: three additional firearms charges against
Cruz, Hernandez, Villega, and Kelly. Counts IV and V vari-
ously charged Hernandez and Cruz with the use of a firearm in

2
The statutory term “cocaine base,” 21 U.S.C.
§ 841(b)(1)(A)(iii), (b)(1)(B)(iii), encompasses but is not lim-
ited to crack cocaine, covering all forms of “cocaine in its
chemically basic form.” DePierre v. United States, 564 U.S.
70, 89 (2011). Because we are concerned here specifically with
crack, however, we will refer simply to it.

9
relation to or in furtherance of a drug-trafficking crime, 18
U.S.C. § 924(c).3 And Count VI charged Cruz, Hernandez, Vil-
lega, and Kelly4 under 18 U.S.C. § 924(o)—conspiracy to vio-
late § 924(c).
B. Jury Selection
One year later, in September 2015, twelve of the
twenty-one defendants proceeded to a consolidated trial before
the Honorable Yvette Kane. On Friday, September 18, with
jury selection set to begin the following Monday, the District
Court issued a series of orders related to the upcoming voir
dire. See D. Ct. Dkt. Nos. 733-40. One such order stated:
AND NOW, on this 18th day of September,
2015, IT IS HEREBY ORDERED THAT due to
courtroom capacity limitations, only (1) court
personnel, (2) defendants, (3) trial counsel and
support staff, and (4) prospective jurors shall be
allowed in the courtroom during jury selection.
No other individuals will be present except by
express authorization of the Court.
App. 10.5 Other than the concern with “courtroom capacity
limitations,” there is no further indication in the record of the

3
The District Court later dismissed Count IV on the Govern-
ment’s motion at the conclusion of its case in chief.
4
Also on the Government’s motion at the end of its case in
chief, the District Court dismissed Count VI as to Villega and
Kelly.
5
All references to the Appendix simpliciter are to three con-
secutively paginated appendices: Volumes I and II of the Her-
nandez Appendix (pages 1-295), the Government’s Supple-
mental Appendix in Rice’s case (pages 296-6902), and the Vil-
lega Appendix (pages 6903-7018).

10
District Court’s rationale for conditionally barring the public
from the jury-selection proceedings. There is also no evidence
of an objection to the order by either the prosecution or the de-
fense, nor is there any evidence of a news organization or other
member of the public either seeking the District Court’s “ex-
press authorization” or being turned away by court officials af-
ter attempting to attend the proceedings.
Jury selection lasted for two days, concluding on Tues-
day, September 22. During the process, Cruz’s trial counsel,
Michael Wiseman, brought a Batson challenge to the Govern-
ment’s first peremptory strike of a prospective juror. The Dis-
trict Court heard the objection in chambers rather than in the
courtroom itself, announcing its decision to do so in open
court. The District Court ultimately ruled that the Govern-
ment’s strike was not motivated by purposeful discrimination.
After the hearing, several defense counsel, led by John Ya-
ninek, counsel for Maurice Atkinson, objected to the District
Court’s decision to hear the challenge out of open court. The
District Court provided a detailed description of the hearing
and the reasons for its ruling, and Yaninek pursued the objec-
tion no further at the time. All defense counsel thereafter pro-
fessed themselves satisfied with the jury members, who were
duly sworn.
The trial commenced the next day, September 23, 2015.
It appears that all other proceedings were open to the public.
C. Trial
The Government’s theory was that the defendants’ iden-
tification with the South Side constituted a continuing, willful
participation in a criminal enterprise. The defense generally
countered that, despite the illegal activity that undoubtedly oc-
curred, expressions of a South Side identity reflected at most a
kind of autochthonous pride, a loyalty borne of a common

11
home, and did not amount to the existence of a South Side gang
or criminal organization.
Witnesses depicted widespread drug trafficking that
was organized, or at least differentiated, according to street
blocks. Each block had a group, or “crew,” of individuals who
would “affiliate with each other,” chiefly through selling
drugs, and in particular crack cocaine. App. 1523. Some crews’
operations were more organized or structured, but a person
from any of the crews could, without incident, sell drugs
throughout the South Side. The most prominent of these groups
was located at Maple and Duke Streets, near what was called
the Jungle—an area formed by four streets, George, Queen,
South, and Maple, with Duke running through it. The Maple
and Duke crew was said to be made up largely of an older gen-
eration of South Side drug dealers. At various points, witnesses
associated Rice, Schueg, Atkinson, Anthony Sistrunk, and
Tyree Eatmon with Maple and Duke, while Williams was said
to be part of another crew, Maple and Manor. By contrast, wit-
nesses described Cruz, Hernandez, and Kelly as principally
distributors of crack to street-level dealers. Villega was identi-
fied as an associate of Cruz and Hernandez who dealt in crack
and heroin.
Together with the descriptions of drug trafficking were
accounts of episodic violence. Members of the crews would
carry or store away firearms for protection, and they would of-
ten retaliate when a fellow South Side member was attacked.
These episodes frequently involved individuals from Parkway,
who were described as rivals, but also occasionally other per-
sons. Witnesses recalled, among other incidents, reprisals for
the wanton killing of a nine-year-old girl, Ciara Savage, on
Mother’s Day in 2009, a violent altercation between South
Side and Parkway members at a gas station and store named
Rutter’s, and the severe beating and eventual murder of a man

12
in the parking lot of a York restaurant called MoMo’s. Such
episodes, the Government charged, were overt acts in further-
ance of the criminal enterprise, reflecting among other things
the preservation of territory and reputation. In general, the de-
fense sought to present these acts of violence as the product of
personal feuds, rather than as indicative of a commitment to a
larger operation.
D. Verdicts and Sentencing
The jury returned its verdicts on November 16, 2015,
announcing them seriatim, with only the relevant defendant
present. All twelve defendants were found guilty on one or
more of the counts against them. They were subsequently sen-
tenced to various periods of incarceration and ordered to pay
certain fines and costs.
The convictions and sentences of imprisonment of the
ten Defendants who have appealed to our Court are as follows:
• Williams: Convicted on Count III; sentenced to 60
months of imprisonment.6
• Cruz: Convicted on Counts I, II, III, V, and VI; sen-
tenced to life terms of imprisonment on Counts I-III,
5 years on Count V, and 20 years on Count VI. The
terms on Counts I-III and VI are concurrent; the term
on Count V is consecutive to those sentences.
• Hernandez: Convicted on Counts I, II, III, V, and VI;
sentenced to life terms of imprisonment on Counts
I-III, 20 years on Count VI, and 60 months on Count
V. The terms on Counts I-III and VI are concurrent;
the term on Count V is consecutive to the other sen-
tences.

6
Williams’s conviction on Count III was for 28-280 grams of
crack cocaine and some marijuana.

13
 • Villega: Convicted on Counts I, II, and III; sen-
tenced to 300 months in prison on each count, to be
served concurrently.
• Rice: Convicted on Counts II and III; sentenced to
200 months in prison on each count, to be served
concurrently.
• Kelly: Convicted on Counts I, II, and III; sentenced
to life terms of imprisonment on each count, to be
served concurrently.
• Schueg: Convicted on Counts II and III; sentenced
to 165 months in prison on each count, to be served
concurrently.
• Atkinson: Convicted on Counts I, II, and III; sen-
tenced to life terms of imprisonment on each count,
to be served concurrently.
• Sistrunk: Convicted on Counts I, II, and III; sen-
tenced to 360 months in prison on each count, to be
served concurrently.
• Eatmon: Convicted on Counts I, II, and III; sen-
tenced to 260 months in prison on each count, to be
served concurrently.
On appeal, these Defendants raise numerous issues, described
above, touching their convictions and sentences.7 We have ju-
risdiction to resolve these issues under 28 U.S.C. § 1291 and

7
While these appeals were pending, on several occasions our
Clerk’s Office encouraged the Defendants to adopt, pursuant
to Federal Rule of Appellate Procedure 28(i), portions of al-
ready-filed briefs rather than raise and argue duplicative issues.
We appreciate that the Defendants followed those suggestions,
but we have also made clear that general statements of adop-
tion under Rule 28(i) will not be regarded. We will not serve

14
18 U.S.C. § 3742(a).8
II. THE PUBLIC-TRIAL ERROR
We begin with the District Court’s closure of the court-
room to the public during jury selection. Because a ruling for
the Defendants on this issue would entail a reversal of their
convictions and remand for a new trial, we confront this ques-
tion at the outset. For the reasons that follow, we will not exer-
cise our discretion to correct the error.
A. Our Review Is for Plain Error
Review of a constitutional error of criminal procedure
is at bottom a matter of rights and remedies: whether a consti-
tutional right has been violated, and whether a remedy shall be
provided for that violation. The District Court’s closure of the
courtroom undoubtedly violated the Defendants’ Sixth
Amendment right to public trial, Presley v. Georgia, 558 U.S.
209, 213 (2010) (per curiam), and under Supreme Court prec-
edent that sort of violation is a “structural” error, see Arizona
v. Fulminante, 499 U.S. 279, 310 (1991) (citing Waller v.
Georgia, 467 U.S. 39, 49 n.9 (1984)). Ordinarily contrasted
with constitutional errors subject to “harmless-error analysis,”
Fulminante, 499 U.S. at 306, this category represents “a lim-
ited class of fundamental constitutional errors that,” Neder v.

as a Defendant’s lawyer, “scour[ing] the record” for him and
determining “which of the many issues of his codefendants
[are] worthy of our consideration.” United States v. Fattah, 914
F.3d 112, 146 n.9 (3d Cir. 2019). We will resolve only those
issues specifically and explicitly identified by each Defendant,
noting where relevant a Rule 28(i) adoption. All else results in
“abandonment and waiver.” Id.
8
The District Court had jurisdiction under 18 U.S.C. § 3231.

15
United States, 527 U.S. 1, 7 (1999), by their very nature, “af-
fect substantial rights” and so cannot be “disregarded,” Fed. R.
Crim. P. 52(a). As a result, in determining the availability of a
remedy, no further inquiry may be necessary beyond the fact
of the violation itself: the injured parties are entitled to “auto-
matic reversal.” Neder, 527 U.S. at 7.
Yet the Federal Rules of Criminal Procedure also dis-
tinguish between preserved and unpreserved errors. A party
can invoke Rule 52(a) on appeal only if he timely objected to
the error, thus giving the district court the opportunity to rec-
tify, or at least respond to, the purported problem. See Fed. R.
Crim. P. 51(b) (describing the procedure for contemporaneous
objection). If the Defendants had done so here, and the District
Court responded inadequately, then they would indeed be en-
titled to a new trial. But they did not object; and regardless of
the nature of the error, in direct appeals from judgments of con-
viction in the federal system, when there is no contemporane-
ous objection in the district court, our review must be for plain
error under Federal Rule of Criminal Procedure 52(b). Johnson
v. United States, 520 U.S. 461, 466 (1997).
A federal appellate court’s authority to remedy an un-
preserved error “is strictly circumscribed.” Puckett v. United
States, 556 U.S. 129, 134 (2009). Following the text of Rule
52(b), the Supreme Court has described a four-part inquiry for
plain-error review. There must: (1) be an “error” that (2) is
“plain” and (3) “affects substantial rights.” United States v.
Olano, 507 U.S. 725, 732 (1993) (alteration omitted) (quoting
Fed. R. Crim. P. 52(b)). If these three conditions are satisfied,
then it is “within the sound discretion of the court of appeals”
to correct the forfeited error—but only if (4) “the error ‘seri-
ously affects the fairness, integrity or public reputation of judi-
cial proceedings.’” Id. (alteration omitted) (quoting United
States v. Young, 470 U.S. 1, 15 (1985)). “Meeting all four

16
prongs is difficult, as it should be.” Puckett, 556 U.S. at 135
(internal quotation marks omitted). In this appeal, the Govern-
ment concedes that the District Court committed an error, and
that the error is plain. The dispute concerns Olano’s third and
fourth prongs.
B. Olano Prong Three
“[I]n most cases,” for an unpreserved error to affect sub-
stantial rights it “must have been prejudicial”—that is, “[i]t
must have affected the outcome of the district court proceed-
ings.” Olano, 507 U.S. at 734. The defendant ordinarily has the
burden of showing “a reasonable probability that, but for the
error claimed, the result of the proceeding would have been
different.” United States v. Dominguez Benitez, 542 U.S. 74,
82 (2004) (alteration omitted) (quoting United States v. Bagley,
473 U.S. 667, 682 (1985) (opinion of Blackmun, J.)). How-
ever, the Court in Olano also acknowledged that “[t]here may
be a special category of forfeited errors that can be corrected
regardless of their effect on the outcome.” 507 U.S. at 735.
Hernandez urges us not only to associate this “special cate-
gory” with structural error, but also to give the error here the
same effect it would have in the Rule 52(a) context—automatic
reversal of the convictions. We cannot accept this argument.
The Supreme Court has never held that Olano’s “special
category” includes or is the same as that of structural error. It
therefore remains at least unclear whether a structural error
ipso facto satisfies Olano’s third prong. The Court has consist-
ently acknowledged but declined to address this possibility.
See United States v. Marcus, 560 U.S. 258, 263 (2010); Puck-
ett, 556 U.S. at 140-41; United States v. Cotton, 535 U.S. 625,
632-33 (2002); Johnson, 520 U.S. at 469; see also Dominguez
Benitez, 542 U.S. at 82 (suggesting that Olano’s third prong
should be treated as “[]tethered to a prejudice requirement” in

17
cases of “nonstructural error”). We too find it unnecessary to
take that doctrinal leap here. Because, as detailed below, a fed-
eral appellate court’s evaluation of Olano’s fourth prong is in-
dependent of whether the third has been satisfied, and the Dis-
trict Court’s error in this case did not “seriously affect the fair-
ness, integrity or public reputation of judicial proceedings,”
Olano, 507 U.S. at 736, we do not need to decide whether the
error also affected the Defendants’ substantial rights.9
C. Olano Prong Four
1. Structural Error Generally
The fact that a type of error has been deemed “struc-
tural” has no independent significance for applying Olano’s
fourth prong. In all direct appeals arising in the federal system,
“the seriousness of the error claimed does not remove consid-
eration of it from the ambit of the Federal Rules of Criminal
Procedure.” Johnson, 520 U.S. at 466. Rule 52(b) states that a
court “may” consider “[a] plain error that affects substantial
rights.” If Olano’s first three prongs are satisfied, the court of
appeals has the “authority” to notice the error, “but is not re-
quired to do so.” Olano, 507 U.S. at 735. “[A] plain error af-
fecting substantial rights does not, without more, satisfy”
Olano’s fourth prong. Id. at 737. Thus, even if we accepted that

9
Our dissenting colleague would presume prejudice given the
nature of the error at issue here. See Dissenting Op. at III.A.
We emphasize that in declining to conduct an inquiry at prong
three, we intend no suggestion that the present error, or any
structural error, does not warrant a presumption of prejudice.
Our conclusion at prong four simply renders a decision on that
question unnecessary, and we will not go out of our way to
make new law. The dissent, by contrast, must address prong
three because of its contrary conclusion at prong four.

18
a structural error necessarily affects substantial rights, our de-
cision would still be an exercise of discretion, calling for an
independent inquiry on the fourth prong.10
Nevertheless, although a structural error is not to be
given automatic effect in the Rule 52(b) context, the same con-
siderations that in other contexts render its correction auto-
matic may coincide with the appropriate exercise of judicial
discretion to notice an unpreserved error. A structural defect is
an error “affecting the framework within which the trial pro-
ceeds, rather than simply an error in the trial process itself.”
Fulminante, 499 U.S. at 310. When such an error occurs over
a contemporaneous objection, the trial “cannot reliably serve
its function as a vehicle for determination of guilt or innocence,
and no criminal punishment may be regarded as fundamentally
fair.” Id. (quoting Rose v. Clark, 478 U.S. 570, 577-78 (1986)).
The origins of Rule 52(b) lie in the recognition that “if a plain
error was committed in a matter so absolutely vital to defend-
ants,” the reviewing court is “at liberty to correct it.” Wiborg
v. United States, 163 U.S. 632, 658 (1896). When the error
threatens “the fair and impartial conduct of the trial,” the fact
that it was not raised contemporaneously “does not preclude
[the appellate court] from correcting [it].” Brasfield v. United
States, 272 U.S. 448, 450 (1926). As the Supreme Court said
in its most recent case on this issue, “the public legitimacy of
our justice system relies on procedures that are neutral, accu-

10
It is true, as Hernandez points out, that our Court has in the
past “assume[d]” in dictum that a structural error “would con-
stitute per se reversible error even under plain error review.”
United States v. Syme, 276 F.3d 131, 155 n.10 (3d Cir. 2002).
Yet we are not bound by that statement, and it is in any event
contrary to the Supreme Court guidance just detailed.

19
rate, consistent, trustworthy, and fair, and that provide oppor-
tunities for error correction.” Rosales-Mireles v. United States,
138 S. Ct. 1897, 1908 (2018) (internal quotation marks omit-
ted).
Therefore, even when confronting a structural error, a
federal court of appeals should evaluate the error in the context
of the unique circumstances of the proceeding as a whole to
determine whether the error warrants remedial action. See id.
at 1909 (“[A]ny exercise of discretion at the fourth prong of
Olano inherently requires ‘a case-specific and fact-intensive’
inquiry.” (quoting Puckett, 556 U.S. at 142)). The very nature
of the error may warrant a remedy in the ordinary case, id. at
1909 n.4, and actual innocence is dispositive, Olano, 507 U.S.
at 736, but these are not the same as automatic reversal. In all
direct appeals from a criminal conviction in the federal system,
the discretion contemplated by Rule 52(b) is to be preserved.
2. Public-Trial Error Specifically
This conclusion receives additional support from our
own and the Supreme Court’s case law on violations of the
Sixth Amendment right to a public trial.
The presence of a contemporaneous objection is an im-
portant reason why violations of that right were deemed struc-
tural error. As early as 1949—in a case, like the present ones,
from the Middle District of Pennsylvania—our Court reversed
a criminal conviction and remanded for a new trial due to a
Sixth Amendment public-trial violation. United States v. Kobli,
172 F.2d 919, 924 (3d Cir. 1949) (en banc). In doing so, we
held “that the Sixth Amendment precludes the general indis-
criminate exclusion of the public from the trial of a criminal
case in a federal court over the objection of the defendant.” Id.
at 923 (emphasis added). Further, in a later case we maintained
that “a defendant who invokes the constitutional guarantee of

20
a public trial need not prove actual prejudice” on appeal.
United States ex rel. Bennett v. Rundle, 419 F.2d 599, 608 (3d
Cir. 1969) (en banc).
The Supreme Court has expressed similar sentiments.
Like Rundle, Waller concerned a Sixth Amendment challenge
to a state trial court’s closure of a suppression hearing. Under
its First Amendment precedent, the Court noted, “the right to
an open trial” is generally, but not absolutely, paramount. Wal-
ler, 467 U.S. at 45 (citing, e.g., Press-Enterprise Co. v. Supe-
rior Court, 464 U.S. 501 (1984)). To justify a closure, there
must be “an overriding interest that is likely to be prejudiced,
the closure must be no broader than necessary to protect that
interest, the trial court must consider reasonable alternatives to
closing the proceeding, and it must make findings adequate to
support the closure.” Id. at 48. Waller extended this framework
to the Sixth Amendment, holding “that under the Sixth Amend-
ment any closure of a suppression hearing over the objections
of the accused must meet the tests set out in Press-Enterprise
and its predecessors.” Id. at 47 (emphasis added). The Court
later applied this standard to a state court’s closure of jury se-
lection to the public. Presley, 558 U.S. at 213. As in Waller,
defense counsel had objected contemporaneously. Id. at 210.
Under these cases, then, a violation of the right to a public trial
is a reversible error when a party lodges a contemporaneous
objection and the trial court fails to articulate the interest be-
hind the closure or to make the appropriate findings.
The Supreme Court’s first consideration of a Sixth
Amendment public-trial violation in the absence of a contem-
poraneous objection came in Weaver v. Massachusetts, 137 S.
Ct. 1899 (2017). Yet that case arose not under Rule 52(b), but
rather in a state collateral proceeding, on a claim of ineffective
assistance of counsel. The Court held that, in this context, the
proper standard to apply is the familiar one under Strickland v.

21
Washington, 466 U.S. 668 (1984). See 137 S. Ct. at 1910-12.
While the Sixth Amendment public-trial right “is important for
fundamental reasons,” the Court explained, “in some cases an
unlawful closure might take place and yet the trial still will be
fundamentally fair from the defendant’s standpoint.” Id. at
1910. This reality underlines the importance of a contempora-
neous objection, which gives the trial court “the chance to cure
the violation either by opening the courtroom or by explaining
the reasons for closure.” Id. at 1912. The Court also noted that
“when state or federal courts adjudicate errors objected to dur-
ing trial and then raised on direct review, the systemic costs of
remedying the error are diminished to some extent.” Id. By
contrast, an ineffective-assistance-of-counsel claim first raised
in postconviction proceedings “‘can function as a way to es-
cape rules of waiver and forfeiture and raise issues not pre-
sented at trial,’ thus undermining the finality of jury verdicts.”
Id. (quoting Harrington v. Richter, 562 U.S. 86, 105 (2011)).
The Court concluded that Weaver had not carried his burden to
show either that he had been prejudiced or that the trial was
rendered fundamentally unfair. Id. at 1913.
Our principal question must be whether and how
Weaver’s analysis in the collateral-review context informs
plain-error review of public-trial violations. The conclusion
that not every public-trial violation results in fundamental un-
fairness supports the particularized inquiry described above.
And while the concern with the finality of judgments might os-
tensibly distinguish Weaver’s context from the present one, it
is nevertheless true that reversal for an error raised for the first
time on direct review carries its own “systemic costs.” The
unique considerations raised by appeal on an unpreserved error
should not be disregarded simply because of the nature of the
error. They may be overcome, but not disregarded. See Puckett,

22
556 U.S. at 135 (“We have repeatedly cautioned that any un-
warranted extension of the authority granted by Rule 52(b)
would disturb the careful balance it strikes between judicial ef-
ficiency and the redress of injustice.” (alteration and internal
quotation marks omitted)); United States v. Atkinson, 297 U.S.
157, 159 (1936) (observing that the practice of not correcting
unpreserved errors is in part “founded upon considerations . . .
of the public interest in bringing litigation to an end after fair
opportunity has been afforded to present all issues of law and
fact”).
In sum, both our own and the Supreme Court’s jurispru-
dence on the Sixth Amendment right to a public trial support
the application here of the “case-specific and fact-intensive in-
quiry” that a federal appellate court is normally to conduct un-
der Olano’s fourth prong. Rosales-Mireles, 138 S. Ct. at 1909
(internal quotation marks omitted).
3. The Legal Standard
Given the relative novelty of a public-trial error re-
viewed under Rule 52(b), our inquiry must look to general
principles discernible in our own and the Supreme Court’s case
law on Olano’s fourth prong and its antecedents. Because
“each case necessarily turns on its own facts,” an appellate
court’s exercise of discretion is properly based on its evalua-
tion of which result would most “promote the ends of justice.”
United States v. Socony-Vacuum Oil Co., 310 U.S. 150, 240
(1940). In conducting this evaluation, the Court has frequently
weighed the costs to the fairness, integrity, and public reputa-
tion of judicial proceedings that would result from allowing the
error to stand with those that would alternatively result from
providing a remedy. We will adopt this standard here.
First, in determining the costs of inaction, the Supreme
Court has focused chiefly upon the error’s effect on the values

23
or interests protected by the violated right. For example, at
stake in Rosales-Mireles—which involved a Sentencing
Guidelines calculation error—was the defendant’s liberty, and
an error “reasonably likely to have resulted in a longer prison
sentence than necessary” sufficiently compromised that inter-
est to advise correction. 138 S. Ct. at 1910. A reasonable citi-
zen, the Court noted, would “bear a rightly diminished view of
the judicial process and its integrity” if the error were allowed
to stand. Id. at 1908 (quoting United States v. Sabillon-Umana,
772 F.3d 1328, 1333-34 (10th Cir. 2014)); see also United
States v. Dahl, 833 F.3d 345, 359 (3d Cir. 2016).
Similarly, in other contexts, the Court has looked to the
error’s effect on the jury’s verdict. In Cotton and Johnson, the
interests underlying the right at issue11 were not so compro-
mised that correction was warranted—in each case, notwith-
standing the error, the evidence supporting conviction was
“overwhelming” and “essentially uncontroverted.” Cotton,
535 U.S. at 633; Johnson, 520 U.S. at 470; see also United
States v. Vazquez, 271 F.3d 93, 105-06 (3d Cir. 2001) (en
banc). Likewise, in Young, the harmful effects of a prosecu-
tor’s inappropriate statements—a violation of his “duty to re-
frain from overzealous conduct,” 470 U.S. at 7—were suffi-
ciently “mitigated,” both by improper statements of defense
counsel and by “overwhelming evidence,” id. at 16-19.
Evaluation of the degree to which an error has compro-
mised the violated right’s underlying values or interests does
not, however, necessarily reduce to a determination of whether

11
A criminal defendant is entitled to “a jury determination that
[he] is guilty of every element of the crime with which he is
charged, beyond a reasonable doubt.” Apprendi v. New Jersey,
530 U.S. 466, 477 (2000) (alteration in original) (quoting
United States v. Gaudin, 515 U.S. 506, 510 (1995)).

24
the error likely altered the outcome of the proceeding. Though
a “court of appeals should no doubt correct a plain forfeited
error that causes the conviction or sentencing of an actually in-
nocent defendant,” the Supreme Court has “never held that a
Rule 52(b) remedy is only warranted in cases of actual inno-
cence.” Olano, 507 U.S. at 736 (emphasis in original); see also
Rosales-Mireles, 138 S. Ct. at 1906. In cases predating Cotton,
Johnson, and Young, for example, the Court held that the error
at issue sufficiently compromised the fairness and impartiality
of the trial that correction was justified. See Brasfield, 272 U.S.
at 450; Clyatt v. United States, 197 U.S. 207, 222 (1905). At
the same time, apart from cases of actual innocence, an altered
outcome does not in itself necessitate correction of the error. In
Rosales-Mireles, the Court allowed that “countervailing fac-
tors [could] satisfy the court of appeals that the fairness, integ-
rity, and public reputation of the proceedings will be preserved
absent correction,” though it did not elaborate on what such
factors might be, concluding only that none existed in the case
before it. 138 S. Ct. at 1909.
Second, against these considerations of the costs of in-
action, the Court has weighed the costs to the fairness, integ-
rity, and public reputation of judicial proceedings that would
alternatively result from noticing the error. In Rosales-Mireles,
the Court noted “the relative ease of correcting the error,” id.
at 1908, commenting that “a remand for resentencing, while
not costless, does not invoke the same difficulties as a remand
for retrial does,” id. (quoting Molina-Martinez v. United States,
136 S. Ct. 1338, 1348-49 (2016)); see also United States v.
Williams, 399 F.3d 450, 456 (2d Cir. 2005) (“A legal system
seeks to protect rights, but it also takes into account the costs
in time, resources, and disruption in the lives of participants . .
. that result when a case must be tried a second time.”). And in
Cotton and Johnson, the Court perceived “[t]he real threat . . .

25
to the fairness, integrity, and public reputation of judicial pro-
ceedings” to be if the error were corrected “despite the over-
whelming and uncontroverted evidence that” the outcome of
the proceeding would have been the same regardless. Cotton,
535 U.S. at 634 (internal quotation marks omitted); see also
Johnson, 520 U.S. at 470.
4. Application and Resolution
Applying this standard, we conclude that the District
Court’s error does not warrant reversal of the Defendants’ con-
victions and remand for a new trial.12
First, the costs of inaction, while not negligible, do not
rise to the level recognized in other cases where a remedy has
been provided. The Sixth Amendment’s public-trial guarantee
is “for the benefit of the accused.” Waller, 467 U.S. at 46 (quot-
ing Gannett Co. v. DePasquale, 443 U.S. 368, 380 (1979)). It
is a means of ensuring the fairness of the trial—“that the pres-
ence of interested spectators may keep [the defendant’s] triers
keenly alive to a sense of their responsibility and to the im-
portance of their functions.” Id.; see also United States v. Lnu,
575 F.3d 298, 305 (3d Cir. 2009) (“The knowledge that every
criminal trial is subject to contemporaneous review in the fo-
rum of public opinion is an effective restraint on possible abuse

12
We acknowledge that one of our sister circuits has reached a
different conclusion. See United States v. Negrón-Sostre, 790
F.3d 295, 306 (1st Cir. 2015). However, that case was decided
prior to Weaver and Rosales-Mireles, and it relied in part upon
circuit precedent that Weaver subsequently abrogated. See id.
(stating that Owens v. United States, 483 F.3d 48 (1st Cir.
2007), “guides our analysis”); see also Lassend v. United
States, 898 F.3d 115, 122 (1st Cir. 2018) (acknowledging
Weaver’s abrogation of Owens).

26
of judicial power.” (quoting Gannett, 443 U.S. at 380)). More
broadly, public access to trial proceedings helps sustain public
confidence that standards of fairness are being observed. See
Press-Enterprise, 464 U.S. at 509.
The District Court’s September 18 order stated that,
“due to courtroom capacity limitations,” only court personnel,
defendants, trial counsel and support staff, and prospective ju-
rors would be allowed in the courtroom during jury selection.
App. 10. All other individuals could be present only “by ex-
press authorization of the Court.” Id. As noted above, the rec-
ord gives no further indication of the District Court’s rationale
for issuing the order. There is no evidence that any party or
member of the press or public objected to the order, nor is there
any evidence of an individual or news organization either seek-
ing authorization from the District Court or being turned away
after attempting to attend the proceedings. Jury selection ulti-
mately lasted only two days, September 21 and 22, with the
trial beginning on September 23. All other proceedings were
open to the public, and a transcript of the jury voir dire was
later made available.
Even on this sparse record, there are facts that suggest
some costs should the error remain uncorrected. The closure
order came from the District Court itself and extended across
an entire phase of the trial. The Court apparently issued the or-
der unprompted, and there is no indication that it—albeit with-
out objection to the order by the parties, counsel, or the pub-
lic—considered reasonable alternatives. It is undeniable that
the order to some degree compromised the values underlying
the public-trial right. It had the potential to call into question
the fairness, integrity, and public reputation of judicial pro-
ceedings because it stamped the violation of the Defendants’

27
Sixth Amendment right with the imprimatur of the federal ju-
diciary itself, thereby undermining public confidence in its im-
partiality.
Nevertheless, there are several countervailing factors
that sufficiently mitigate this possibility. For one, although the
closure encompassed all of the jury-selection phase, those pro-
ceedings lasted only two days; the public had access to all other
phases of the trial, which in total lasted longer than seven
weeks. See, e.g., Weaver, 137 S. Ct. at 1913 (“The closure was
limited to the jury voir dire; the courtroom remained open dur-
ing the evidentiary phase of the trial.”); Press-Enterprise, 464
U.S. at 510 (finding it significant that “[a]lthough three days of
voir dire in this case were open to the public, six weeks of the
proceedings were closed” (emphasis in original)). Further, a
transcript of the proceedings was produced and later disclosed.
See D. Ct. Dkt. Nos. 974-993, 997-1005, 1024-1027; see also
Weaver, 137 S. Ct. at 1913; Press-Enterprise, 464 U.S. at 513.
And as our Court has said, “[i]t is access to the content of the
proceeding—whether in person, or via some form of documen-
tation—that matters.” United States v. Antar, 38 F.3d 1348,
1359-60 (3d Cir. 1994) (emphasis omitted).13 Moreover,
knowledge both of the media’s attention to the trial and of the
transcript’s production (which ensures publicity in perpetuity)
may have had a similar effect on the proceedings’ participants
as real-time public access would have had, keeping them

13
This is not to suggest, as Antar makes clear, that subsequent
release of the transcript may substitute for closure. See 38 F.3d
at 1360 n.13. Our point here is that, for purposes of plain-error
review, subsequent disclosure of the transcript, while not a per-
fect substitute, at least mitigates the harm caused by the clo-
sure.

28
“keenly alive to a sense of their responsibility and to the im-
portance of their functions.” Waller, 467 U.S. at 46 (quoting
Gannett, 443 U.S. at 380). In addition, although the general
public was not, absent authorization, able to be present at jury
selection, as in Weaver, “there were many members of the ve-
nire who did not become jurors but who did observe the pro-
ceedings.” 137 S. Ct. at 1913. Finally, there has been “no sug-
gestion of misbehavior by the prosecutor, judge, or any other
party; and no suggestion that any of the participants failed to
approach their duties with the neutrality and serious purpose
that our system demands.” Id.
The ways, then, in which the closure potentially com-
promised the values protected by the Defendants’ Sixth
Amendment right are answered by countervailing factors sug-
gesting that those values were in other respects substantially
vindicated—that, in spite of the closure, the jury-selection pro-
ceedings possessed the publicity, neutrality, and professional-
ism that are essential components of upholding an accused’s
right to a fair and public trial. Allowing the error to stand would
not leave in place an unmitigated nullification of the values and
interests underlying the right at issue.
Second, the costs of remedial action here would be sig-
nificant. Unlike in Rosales-Mireles, we are confronted with a
remand for a new trial in ten consolidated cases whose original
trial occurred almost five years ago, spanned approximately
two months, and involved well over one hundred witnesses.
But even in the absence of the heavy burdens specific to these
cases, the prospect of retrial demands “a high degree of cau-
tion,” Rosales-Mireles, 138 S. Ct. at 1909, and implicates more
fully the Supreme Court’s admonition that we exercise our dis-
cretion under Rule 52(b) “sparingly,” id. (quoting Jones v.
United States, 527 U.S. 373, 389 (1999)). Moreover, when the

29
Supreme Court in Waller acknowledged a public-trial error un-
der the Sixth Amendment, it did not automatically reverse the
convictions and remand for a new trial. Even there, on review
of a preserved error, it cautioned that “the remedy should be
appropriate to the violation” and contemplated the possibility
that in some instances “a new trial . . . would be a windfall for
the defendant, and not in the public interest.” 467 U.S. at 50.
The same general consideration applies here: the remedy is to
be assessed relative to the costs of the error.
***
The practical costs of correcting the District Court’s er-
ror are not dispositive,14 but when we consider them along with
the mitigated costs of inaction, we decline to exercise our dis-
cretion in this instance. The importance of the “searchlight” of
the public trial is “deeply rooted” in the history of our federal
constitutional order and system of justice; and it has long been
a feature of our Court’s jurisprudence. Rundle, 419 F.2d at 605-
06. Nevertheless, on this record, we cannot say that the values
underlying the Defendants’ right to a public trial were suffi-
ciently compromised that the costs to the fairness, integrity,
and public reputation of judicial proceedings that would result

14
There was some dispute at oral argument over the analytical
significance of sandbagging, despite no suggestion that it oc-
curred here. See Oral Arg. at 2:53:28-2:55:54; 3:01:24-
3:02:30. Although sandbagging can be a concern, see United
States v. Bansal, 663 F.3d 634, 661 (3d Cir. 2011), we decline
here to give it weight. For one, it is already accounted for doc-
trinally through the Olano test. See Puckett, 556 U.S. at 134.
And the specter of sandbagging is most acute where the prece-
dent established would be an automatic new trial. Under our
standard, there is no such automaticity, each case turning on its
own facts.

30
from letting the District Court’s error stand outweigh those that
would alternatively result from reversing the Defendants’ con-
victions and remanding for a new trial. We cannot, in sum, say
that the District Court’s closure of jury selection to the public
“seriously affect[ed] the fairness, integrity, or public reputation
of judicial proceedings.” Olano, 507 U.S. at 736.15
III. RIGHT-TO-PRESENCE CHALLENGE
Atkinson argues that the District Court’s in camera res-
olution of the Batson challenge during jury selection violated
his constitutional “right to personal presence at all critical
stages of the trial.” Rushen v. Spain, 464 U.S. 114, 117 (1983)
(per curiam); see also Kentucky v. Stincer, 482 U.S. 730, 745
(1987). He further contends that the exclusion was sufficiently
prejudicial to warrant a new trial. The Supreme Court has made

15
Our dissenting colleague places great weight on the distinc-
tion between harmless and structural error. He suggests that in
considering the costs of letting the error stand, we improperly
“rel[y] on cases that consider errors reviewed for harmless-
ness.” Dissenting Op. at III.B. And rather than accounting for
the costs of correction, he thinks “[t]he nature of the error . . .
must be the lodestar of our” analysis. Id. But “the term ‘struc-
tural error’ carries with it no talismanic significance as a doc-
trinal matter.” Weaver, 137 S. Ct. at 1910. The present cases
ask us to weigh the intersection of two fundamental distinc-
tions in criminal procedure: harmless and structural error, and
preserved and unpreserved error. The dissent would give dis-
positive weight to the former. In our view, at least in the con-
text of public-trial errors, neither the case law nor the compet-
ing values at stake warrant that approach. And to the extent the
dissent simply weighs the costs of inaction differently here, we
acknowledge his concerns, but respectfully reach the opposite
conclusion on the facts before us.

31
clear that violations of the right to be present are subject to
harmless-error review. See Fulminante, 499 U.S. at 306-07
(citing Rushen, 464 U.S. at 117-18 & n.2). We may assume
without deciding that there was a violation here, because even
if an error occurred, “it was harmless beyond a reasonable
doubt.” Chapman v. California, 386 U.S. 18, 24 (1967).16
In evaluating a putative equal protection violation under
Batson, trial courts are to follow a three-step process.
First, a defendant must make a prima facie show-
ing that a peremptory challenge has been exer-
cised on the basis of race. Second, if that show-
ing has been made, the prosecution must offer a
race-neutral basis for striking the juror in ques-
tion. Third, in light of the parties’ submissions,
the trial court must determine whether the de-
fendant has shown purposeful discrimination.
Miller-El v. Cockrell, 537 U.S. 322, 328-29 (2003) (citing Bat-
son, 476 U.S. at 96-98). “[T]he job of enforcing Batson rests
first and foremost with trial judges,” who may consider a num-
ber of factors in determining whether racial discrimination has
occurred. Flowers v. Mississippi, 139 S. Ct. 2228, 2243 (2019).
These include: whether the prosecutor’s proffered explanations
are pretextual, see Snyder v. Louisiana, 552 U.S. 472, 485
(2008), which can be shown through “side-by-side compari-
sons of some black venire panelists who were struck and white
panelists allowed to serve,” Miller-El v. Dretke, 545 U.S. 231,
241 (2005); “a prosecutor’s misrepresentations of the record
when defending the strike[],” Flowers, 139 S. Ct. at 2243; and
any other “circumstantial evidence that ‘bears upon the issue
of racial animosity,’” Foster v. Chatman, 136 S. Ct. 1737, 1754

16
Kelly adopts Atkinson’s argument under Rule 28(i).

32
(2016) (alteration omitted) (quoting Snyder, 552 U.S. at 478)).
Batson’s third step “turns on factual determinations, and, ‘in
the absence of exceptional circumstances,’ we defer to [trial]
court factual findings unless we conclude that they are clearly
erroneous.” Id. at 1747 (quoting Snyder, 552 U.S. at 477)).
Here, there is no reasonable basis for concluding that
prejudice resulted from the District Court’s conduct of the Bat-
son hearing. At no point during the hearing or afterward did the
District Court or defense counsel suggest that any of the Gov-
ernment’s proffered reasons were pretextual, that the Govern-
ment had misrepresented the record, or that any other circum-
stantial evidence suggested racial bias. Indeed, Wiseman—
who had raised the objection and was one of two defense coun-
sel present—acknowledged at the hearing, and Atkinson con-
cedes on appeal, that the Government “stated race-neutral rea-
sons.” App. 667. And when Wiseman and Royce Morris, the
other defense attorney present, questioned whether the charac-
teristics that led the Government to strike the juror were unique
among the persons in the venire, the District Court proceeded,
with Wiseman and Morris’s assistance, to search the question-
naires for any other remaining juror with characteristics similar
those for which the juror was struck—in particular, the exist-
ence of multiple relatives who had been criminally convicted
and imprisoned, including for drug trafficking. The search re-
vealed no comparable jurors still on the panel. The record be-
fore us provides no basis for doubting the District Court’s side-
by-side comparison of the jurors. See Davis v. Ayala, 576 U.S.
257, 274 (2015).17 Finally, we have not been shown any evi-

17
Ayala was decided under the stricter standard applied on ha-
beas review of a state court decision. See 576 U.S. at 267-68.
However, the Court gives no indication that its decision on this

33
dence that might otherwise contradict the Government’s repre-
sentations or suggest that it acted on grounds of racial animus.
In sum, we have no reason to conclude that Atkinson’s
absence from the Batson hearing was prejudicial. If, therefore,
“the alleged constitutional error” occurred, it was “harmless
beyond a reasonable doubt.” Rushen, 464 U.S. at 121.
IV. EVIDENTIARY CHALLENGES
The Defendants’ evidentiary challenges fall into three
basic categories. First, Kelly and Sistrunk appeal the District
Court’s denial of their motions to suppress evidence obtained
from searches of their residences. Second, Atkinson asserts
that the Government knowingly persisted in the use of perjured
testimony, thus violating his constitutional right to due process.
Finally, those Defendants and four others challenge some of
the District Court’s decisions regarding the admission of evi-
dence. We find no error in any instance.
A. Suppression
Shortly after the grand jury returned its initial indict-
ment in March 2014, federal agents searched Kelly’s apartment
at 337 East Philadelphia Street in York, seizing evidence later
introduced at trial. Almost exactly six months later, just after
the return of the second superseding indictment, agents con-
ducted a similar search of Sistrunk’s apartment, located at 326
West Philadelphia Street, also seizing evidence that was later
introduced. The Government conducted each search pursuant
to a warrant issued by Magistrate Judge Carlson. ATF Special
Agent Scott Endy signed the warrant applications and attached
a sworn affidavit to each of them, detailing his decades-long
experience in federal law enforcement, the history of the South

point would have been different under the “clear error” stand-
ard we are to apply here. See Foster, 136 S. Ct. at 1747.

34
Side investigation, and the basis for probable cause. To estab-
lish the latter, he relied in part upon information provided by
several confidential informants relating to Kelly and Sistrunk’s
drug-trafficking activities.
Approximately two months before the trial, Kelly and
Sistrunk filed motions to suppress the evidence obtained from
the searches. They contended that the information in the affi-
davits was insufficient to establish a factual basis for probable
cause and that the exclusionary rule’s good-faith exception did
not apply. The District Court held hearings on the motions on
August 28, 2015 and denied both of them less than a week later.
It included with each of its orders a memorandum explaining
its decision. Kelly and Sistrunk now appeal those orders, rais-
ing largely the same arguments they did before the District
Court.
1. Kelly
“[N]o Warrants shall issue,” the Fourth Amendment de-
clares, “but upon probable cause, supported by Oath or affir-
mation, and particularly describing the place to be searched,
and the persons or things to be seized.” U.S. Const. amend. IV.
This clause was intended “to affirm and preserve a cherished
rule of the common law, designed to prevent the issue of
groundless warrants.” McGrain v. Daugherty, 273 U.S. 135,
156 (1927). We are satisfied that the warrant to search Kelly’s
residence was not groundless: Special Agent Endy’s affidavit
supplied a sufficient basis for probable cause.
The Legal Standard
“Our review of the denial of a motion to suppress is for
clear error as to the District Court’s findings of fact, and ple-
nary as to legal conclusions in light of those facts.” United
States v. Hester, 910 F.3d 78, 84 (3d Cir. 2018). In contexts
like the present, though, that latter standard applies only to our

35
review of “the District Court’s evaluation of the magistrate’s
probable cause determination.” United States v. Stearn, 597
F.3d 540, 554 (3d Cir. 2010). We pay great deference to the
magistrate’s initial determination, asking only “whether ‘the
magistrate had a substantial basis for concluding that probable
cause existed.’” Id. (quoting Illinois v. Gates, 462 U.S. 213,
238-39 (1983)). It is distinctly the magistrate’s task to make
the “practical, common-sense decision whether, given all the
circumstances set forth in the affidavit before him, including
the veracity and basis of knowledge of persons supplying hear-
say information, there is a fair probability that contraband or
evidence of a crime will be found in a particular place.” Gates,
462 U.S. at 238 (internal quotation marks omitted).
Specifically, “[w]hen the crime under investigation is
drug distribution, a magistrate may find probable cause to
search the target’s residence even without direct evidence that
contraband will be found there.” Stearn, 597 F.3d at 558. We
have long maintained that when a suspect is involved in drug
trafficking, on a significant scale or for an extended period of
time, it is reasonable to infer that he would store evidence of
that illicit activity in his home. See United States v. Hodge, 246
F.3d 301, 306 (3d Cir. 2001); United States v. Whitner, 219
F.3d 289, 297-98 (3d Cir. 2000). It is insufficient, however, if
the affidavit suggests only that the suspect “is actually a drug
dealer” and “that the place to be searched is possessed by, or
the domicile of, the [suspect].” United States v. Burton, 288
F.3d 91, 104 (3d Cir. 2002). There must also be evidence “link-
ing [the targeted location] to the [suspect]’s drug activities.”
Id. (emphasis added). “[T]he search of a drug dealer’s home
would be unreasonable if the affidavit suggested no reason to
believe contraband would be found there.” Stearn, 597 F.3d at
559.

36
 Further, when (as here) the affidavit refers to infor-
mation gained from confidential informants, bare conclusory
assertions by the affiant of the reliability and veracity of the
informants are insufficient. See Gates, 462 U.S. at 239. “Mere
affirmance of belief or suspicion is not enough.” Nathanson v.
United States, 290 U.S. 41, 47 (1933). But when “independent
police work” substantially corroborates the information of a
confidential informant, “an entirely different case” is pre-
sented. Gates, 462 U.S. at 241-42. “[C]orroborat[ion] in sig-
nificant part by independent police investigation” may provide
the requisite substantial basis for a magistrate’s finding of
probable cause, to which we will defer. Stearn, 597 F.3d at 556,
557-58; see also Gates, 462 U.S. at 246.
Application and Resolution
Informants told law enforcement of several interactions
with Kelly related to drug trafficking. In September 2013, an
informant identified Kelly in a photograph and stated that he
had supplied the informant with crack “on numerous occasions
in the recent past.” Kelly App. 120, ¶ 18. Another informant
described a February 2014 encounter in which the informant
asked Kelly for crack to distribute, and Kelly responded that
he was going to Atlantic City to get some more cocaine.
Around that same time, a third informant told a York police
detective that Hernandez was supplying Kelly with large
amounts of crack. These data points suggest that Kelly was at
least involved in the sale and supply of crack cocaine shortly
before the warrant issued.
That suggestion was corroborated by independent po-
lice work. The affidavit describes two incidents that occurred
in September 2013. York law enforcement conducted a con-
trolled delivery of $120 to Kelly through a confidential source
who had been fronted cocaine. Six days later, law enforcement

37
oversaw a controlled buy and delivery of crack involving
Kelly. The source received the drugs earlier in the day, and
later delivered $150 to Kelly “at 337 E. Philadelphia Street.”
Kelly App. 129, ¶ 57. There was some dispute over this word-
ing at the suppression hearing, and Kelly contends on appeal
that it incorrectly implies that the transaction took place inside
his residence, when the police report states that the transaction
occurred in front of the building. For the reasons given above,
however, that distinction is not decisive. The incident at least
indicates that in the months prior to the warrant application,
Kelly was conducting drug transactions in close physical prox-
imity to his apartment.
The final relevant incident in the affidavit is the most
significant. In early March 2014, about two weeks before Kelly
was indicted, federal and local law enforcement (including
Special Agent Endy) conducted a controlled purchase of crack
from Kelly through a cooperating source. Surveillance docu-
mented Kelly leaving his East Philadelphia Street apartment,
driving to the location, delivering (what was later confirmed to
be) crack to the source, and then returning immediately to his
apartment. “While we generally accept the common sense
proposition that drug dealers often keep evidence of their trans-
actions at home, that inference is much stronger when the home
is the first place a drug dealer proceeds following such a trans-
action.” Burton, 288 F.3d at 104 (citation omitted).
In sum, independent police work corroborated the sug-
gestion of multiple informants that Kelly was not an occasional
street-level dealer, but one who consistently sold and supplied
crack to others in the months and weeks leading up to the war-
rant application. Further, that police work provided evidence
placing Kelly’s residence on East Philadelphia Street in close
spatial and temporal proximity to his illegal activity. Magis-
trate Judge Carlson therefore had ample basis to conclude there

38
was “a fair probability that contraband or evidence of a crime
w[ould] be found” at the apartment. Gates, 462 U.S. at 238.
2. Sistrunk
Our Court has “turn[ed] directly to the good faith issue”
when we concluded that a defendant’s probable-cause argu-
ments did not “involve novel questions of law whose resolution
is necessary to guide future action by law enforcement officers
and magistrates.” United States v. Ninety-Two Thousand Four
Hundred Twenty-Two Dollars & Fifty-Seven Cents, 307 F.3d
137, 145 (3d Cir. 2002) (alterations and internal quotation
marks omitted); see United States v. Leon, 468 U.S. 897, 925
(1984). We think such a move is appropriate here, and we will
affirm the denial of Sistrunk’s motion on good-faith grounds.
The Legal Standard
“To trigger the exclusionary rule, police conduct must
be sufficiently deliberate that exclusion can meaningfully deter
it, and sufficiently culpable that such deterrence is worth the
price paid by the justice system.” Herring v. United States, 555
U.S. 135, 144 (2009). One triggering circumstance is when
“the magistrate or judge in issuing a warrant was misled by
information in an affidavit that the affiant knew was false or
would have known was false except for his reckless disregard
of the truth.” Leon, 468 U.S. at 923 (citing Franks v. Delaware,
438 U.S. 154 (1978)). The Franks rule, we have said, encom-
passes not only an affiant’s assertions, but also his omissions.
See Wilson v. Russo, 212 F.3d 781, 787 (3d Cir. 2000). Our
standard for assertions “is that . . . ‘when viewing all the evi-
dence, the affiant must have entertained serious doubts as to
the truth of his statements or had obvious reasons to doubt the
accuracy of the information he reported.’” United States v.
Brown, 631 F.3d 638, 645 (3d Cir. 2011) (quoting Wilson, 212
F.3d at 788). For omissions, by contrast, we ask whether the

39
“officer withholds a fact in his ken that any reasonable person
would have known . . . was the kind of thing the judge would
wish to know.” Wilson, 212 F.3d at 788 (alteration and internal
quotation marks omitted).
Although Wilson concerned an action under 42 U.S.C.
§ 1983, we have also applied it to resolve appeals of judgments
following Franks hearings. See Brown, 631 F.3d at 648-49;
United States v. Yusuf, 461 F.3d 374, 383-84 (3d Cir. 2006).
We will extend this approach to cases where, as here, Franks
is raised in the good-faith context—where the question is only
whether the exclusionary rule should apply. Yet our concern is
with only the first prong of the Franks test—that the affiant
acted deliberately to conceal the truth or with “reckless disre-
gard for the truth.” Franks, 438 U.S. at 171; see Leon, 468 U.S.
at 923. The inquiry at the second prong—that the “false state-
ments or omissions . . . [be] material, or necessary, to the find-
ing of probable cause,” Sherwood v. Mulvihill, 113 F.3d 396,
399 (3d Cir. 1997)—is unnecessary because the presumption
is that a Fourth Amendment violation has occurred. See Her-
ring, 555 U.S. at 145; Leon, 468 U.S. at 922 n.23.
This accordingly demands adjusting the application of
the first prong when an affiant’s alleged omissions are at issue.
In the § 1983 context, we have applied the first prong in light
of the second, asking at the former whether the omitted facts
and circumstances were “relevant to the existence of probable
cause.” Dempsey v. Bucknell Univ., 834 F.3d 457, 471 & n.9
(3d Cir. 2016). But, when good faith is concerned, the proper
question is not simply whether the allegedly omitted infor-
mation was known to the affiant and relevant to the magis-
trate’s probable-cause inquiry, but also whether the deliberate
or reckless omission, if it occurred, was “so objectively culpa-
ble as to require exclusion.” Herring, 555 U.S. at 146; see also

40
Dempsey, 834 F.3d at 473 n.13 (noting that satisfaction of its
standard does not necessarily amount to a finding of bad faith).
Application and Resolution
Sistrunk identifies four instances where Special Agent
Endy allegedly omitted relevant facts, thereby “misle[ading]
the magistrate judge in reckless disregard for the truth.” Sis-
trunk Br. at 26.
First, the affidavit states that on July 8, 2007, “a
Southside gang member” was “fatally shot multiple times.”
Sistrunk App. 170. A suspect later made “a statement to police
[that] implicated Anthony Sistrunk as being . . . with him dur-
ing the shooting.” Id. Sistrunk contends that this statement
“fail[ed] to inform the . . . magistrate that [the suspect] exoner-
ated [him] of any role in th[e] shooting.” Sistrunk Br. at 25.
Second, the affidavit relates that in April 2009, Sistrunk
fled a vehicle stop and was later arrested. Police discovered
two firearms in the vehicle. Sistrunk was later “convicted of
fleeing or attempting to elude police.” Sistrunk App. 170. He
now contends that this account omits the fact that some fire-
arms-related charges were withdrawn, and that the jury acquit-
ted him of other offenses.
Third, according to the affidavit, while Sistrunk was in
prison in September 2009, an ATF Special Agent “obtained the
inmate visitor list for Sistrunk which indicated an association
with multiple Southside Gang members.” Sistrunk App. 170.
Sistrunk argues that this information “failed to report that none
of [his] co-defendants listed on his prison visitor list actually
visited [him].” Sistrunk Br. at 26.
The fourth instance concerns the homicide of Christen
Latham in November 2012. The affidavit states that “police
identified . . . Sistrunk as being involved in an altercation with

41
the victim prior to his murder.” Sistrunk App. 171. This ac-
count, Sistrunk says, omitted that no one was criminally
charged for the homicide, that he was not suspected for the
crime, and that a witness did not identify him as being present.
These alleged omissions do not amount to a deliberate
or reckless concealment of facts both relevant to the magis-
trate’s probable-cause inquiry and evincing a culpability worth
the costs of suppression.18 The context is important. Special
Agent Endy filed his warrant application on September 22,
2014—only five days after the grand jury returned the second
superseding indictment. The application “clearly was sup-
ported by much more than a ‘bare bones’ affidavit”—it “re-
lated the results of an extensive investigation” that had already
led to Sistrunk’s indictment on conspiracy and drug-trafficking
charges. Leon, 468 U.S. at 926. Moreover, none of the suppos-
edly omitted facts negates, or even substantially mitigates, the
intended implication of the related facts actually adduced: that,
as the affidavit asserted, Sistrunk “ha[d] a long history of mem-
bership in the Southside Gang and ha[d] consistently engaged
in or ha[d] been associated with criminal activity including
drug trafficking, firearm possession and violence.” Sistrunk
App. 174. As a result, Special Agent Endy’s failure to include
the facts does not evince the level of culpability necessary to

18
In Brown—which concerned Franks prong one—we held
that the standard of review for assertions is clear error, reason-
ing that a district court’s requisite determination is “essentially
factual.” See 631 F.3d at 642, 644-45. The parties here have
not briefed us on the appropriate standard of review in the
omissions context, and we find it unnecessary to resolve that
question. Even if our review was de novo, we would still affirm
the District Court’s judgment.

42
trigger the exclusionary rule. The costs of suppression here
would far outweigh any concomitant deterrence effect.
B. Knowing Use of Perjured Testimony
During his testimony, Darvin Allen, one of the Govern-
ment’s principal witnesses, described a March 2009 episode of
attack and retaliation between members of South Side and
Parkway. Late one night at a club, Jahkeem Abney, a South
Side member, got into a verbal dispute with some men from
Parkway and was later shot in front of the club. A few days
later, Allen recounted, several persons, including Atkinson,
discussed how to respond to the shooting. Allen then testified
that these same individuals drove up to Parkway and “engaged
in gunfire” with Skylar Handy, one of the Parkway members
at the club the night Abney was shot. App. 1647. On cross-
examination, however, Atkinson’s counsel, Yaninek, asked
Allen if it would “make sense to [him]” that Atkinson was in-
carcerated in March 2009. App. 1801. Allen answered affirm-
atively and agreed that, as a result, Atkinson could not have
been involved in the retaliatory shooting.19 Later, during the
defense portion of the trial, Yaninek questioned Special Agent
Endy, who had prepared Allen for trial. Endy acknowledged
that his report of investigation included Allen’s identification
of Atkinson at the retaliatory shooting, and he accepted that

19
Yaninek had earlier, at a sidebar conversation during direct
examination, moved for a mistrial on the basis of the inaccu-
racy. (Though he mistakenly said the testimony placed Atkin-
son at the club in possession of a gun, rather than simply at the
retaliation.) The District Court denied the motion, declaring the
issue “the proper subject of cross-examination” and not
“grounds for a mistrial.” App. 1664. Atkinson does not appeal
the District Court’s decision to allow the error to be resolved
on cross.

43
this was impossible, but he did not recall Allen testifying to
that effect.
Atkinson now asks for a new trial, contending that the
Government knew of Allen’s error and chose not to correct it.
The Supreme Court has long maintained that under the Due
Process Clauses, the prosecution may neither present nor with-
hold known false evidence, nor “allow[] [such evidence] to go
uncorrected when it appears.” Giglio v. United States, 405 U.S.
150, 153 (1972) (quoting Napue v. Illinois, 360 U.S. 264, 269
(1959)) (citing Brady v. Maryland, 373 U.S. 83, 87 (1963); and
Mooney v. Holohan, 294 U.S. 103, 112 (1935)). Yet such a vi-
olation, if established, does not alone warrant a new trial; there
must also be prejudice (or materiality). See Giglio, 405 U.S. at
154 (citing Napue, 360 U.S. at 271); see also Turner v. United
States, 137 S. Ct. 1885, 1893 (2017). Accordingly, in cases of
uncorrected false testimony, our Court requires a defendant to
show four elements: (1) the witness committed perjury; (2) the
government knew or should have known of the perjury; (3) the
testimony went uncorrected; and (4) there is a reasonable like-
lihood the false testimony affected the verdict. See Lambert v.
Blackwell, 387 F.3d 210, 242 (3d Cir. 2004). Atkinson’s chal-
lenge fails at the first prong.
“A witness commits perjury if he or she ‘gives false tes-
timony concerning a material matter with the willful intent to
provide false testimony, rather than as a result of confusion,
mistake, or faulty memory.’” United States v. Hoffecker, 530
F.3d 137, 183 (3d Cir. 2008) (quoting United States v. Dunni-
gan, 507 U.S. 87, 94 (1993)). Allen’s testimony was not lim-
ited to the night club incident; it ranged across several years
and recounted multiple shootings involving a number of differ-
ent persons. That Allen could not remember precisely who was
present at the March 2009 retaliatory shooting is therefore un-
surprising, and it does not in itself demonstrate willful intent.

44
Further, Atkinson presents no evidence that Allen, at the time
of his direct testimony, knew that Atkinson was incarcerated in
March 2009. Compare Haskell v. Superintendent Greene SCI,
866 F.3d 139, 143, 146 (3d Cir. 2017), with Hoffecker, 530
F.3d at 183. Indeed, on cross-examination, when asked
whether Atkinson was present at the retaliatory shooting, Allen
replied that he knew Atkinson “committed a shooting at Sky-
lar” Handy, but that he didn’t “know if it was March because I
think [Atkinson] went away.” App. 1801. And when Allen was
affirmatively presented with the fact of Atkinson’s incarcera-
tion, he readily allowed it. Given this testimony, we cannot but
conclude that Allen’s initial identification of Atkinson was
simply the result of a “faulty memory.” Hoffecker, 530 F.3d at
183.
C. Admission
The final category of evidentiary challenges concerns
the admission and exclusion of evidence at trial. On multiple
occasions, it is argued, the District Court ran afoul of the rele-
vance provisions of the Federal Rules of Evidence by admitting
evidence that either was unfairly prejudicial in excess of its
probative value or served only to prove a Defendant’s charac-
ter. Several Defendants also challenge the District Court’s ad-
missions decisions regarding expert testimony. We perceive no
error in any of these instances.
1. Relevance
We will disturb a district court’s admission decision
only if the court abused its discretion—if the decision “was ar-
bitrary, fanciful or clearly unreasonable,” such that “no reason-
able person would adopt the district court’s view.” United
States v. Starnes, 583 F.3d 196, 214 (3d Cir. 2009) (internal
citation omitted).

45
 Kelly’s Nickname
The second superseding indictment included an alias, or
street name, for each defendant. The one for Kelly was
“Killer.” App. 18. Early in the trial, his attorney filed a motion
in limine objecting to the Government’s use of the alias as un-
fairly prejudicial because it suggested extrinsic evidence that
Kelly had committed murder. The Government countered that
certain witnesses knew Kelly only through his alias, and that it
would use the nickname only to identify Kelly, thus preventing
jury confusion. The District Court agreed with the Govern-
ment. It also, at the conclusion of the trial, included a limiting
instruction to the jury on this issue. Kelly now seeks a new trial,
arguing that the “probative value” of the nickname evidence
was “substantially outweighed by a danger of . . . unfair preju-
dice.” Fed. R. Evid. 403.
Several of our sister circuits have long maintained that
the prosecution’s use of a defendant’s alias in an indictment or
at trial is permissible where the evidence is relevant—includ-
ing for purposes of identifying the defendant—and does not
result in unfair prejudice. See, e.g., United States v. Doe, 741
F.3d 217, 227 (1st Cir. 2013); United States v. Farmer, 583
F.3d 131, 144-47 (2d Cir. 2009); United States v. Emueg-
bunam, 268 F.3d 377, 394 (6th Cir. 2001); United States v.
Delpit, 94 F.3d 1134, 1146 (8th Cir. 1996); United States v.
Hines, 955 F.2d 1449, 1454 (11th Cir. 1992); United States v.
Williams, 739 F.2d 297, 299-300 (7th Cir. 1984). We agree,
and adopt this standard here.
The District Court’s judgment easily passes muster. Al-
len knew Kelly only by his nickname, and the District Court
engaged in a reasonable balancing of the testimony’s relevance
with the nickname’s potential to generate unfair prejudice.
Kelly points to no instance where either Allen or a later witness

46
in the same position was able to identify him by anything else,
nor does he indicate any moment where the Government used
the alias to do anything other than identify him in a witness’s
testimony.20 Further, the District Court fortified its Rule 403
balancing by including the limiting instruction. We perceive no
abuse of discretion in this course of events.
The Latham Homicide
A few hours after midnight on November 17, 2012, a
Harrisburg man named Christen Latham died of a gunshot
wound to the chest in the parking lot outside a York restaurant
known as MoMo’s. A verbal dispute inside the restaurant
spilled out into the parking lot, where Latham was at first se-
verely beaten by several men and then fatally shot. Police later
identified Hernandez, Cruz, Kelly, and Schueg as either in-
volved in or at least present at the altercation,21 but no charges
were ever filed.
The Government sought at trial to introduce evidence
suggesting the involvement of several defendants in the alter-
cation, including testimony that Hernandez threw the first
punch and circumstantial evidence that Kelly was the one who
killed Latham. Hernandez filed a joint motion in limine to ex-
clude all the evidence, arguing that it was inadmissible under
Federal Rules of Evidence 402, 403, and 404(b). The District

20
Kelly asserts that the Government “prompted” Cordaress
Rogers to use the nickname although Rogers clearly knew
Kelly’s given name. Kelly Reply Br. at 4. We do not read the
testimony that way. It is clear from the context that the Gov-
ernment was seeking to elicit Kelly’s surname, and not his
nickname.
21
Rogers testified at trial that Sistrunk told him that he also
was present.

47
Court denied the motion, ruling that the evidence was intrinsic
to the RICO-conspiracy offense charged at Count I and that
any danger of unfair prejudice did not substantially outweigh
the evidence’s probative value. Seven Defendants22 now con-
test one or both aspects of that ruling.
Intrinsic evidence need not be analyzed under Rule
404(b) because it is not “[e]vidence of any crime, wrong, or
other act,” Fed. R. Evid. 404(b)(1), but rather “part and parcel
of the charged offense,” United States v. Green, 617 F.3d 233,
245 (3d Cir. 2010). We have, however, limited “the ‘intrinsic’
label [to] two narrow categories of evidence”: (1) where the
uncharged conduct “directly proves the charged offense”; and
(2) where it is “performed contemporaneously with the
charged crime” and “facilitate[s] the commission of the
charged crime.” Id. at 248-49 (internal quotation marks omit-
ted). This suggests that the nature and scope of the evidence
able to be deemed intrinsic will vary with the charged offense.
In particular, where a criminal conspiracy is charged, courts
have afforded the prosecution considerable leeway to present
evidence, even of unalleged acts within the indictment period,
that reflects a conspiratorial agreement or furtherance of the
conspiracy’s illegal objectives. See, e.g., United States v. Bush,
944 F.3d 189, 196-97 (4th Cir. 2019); United States v. McGill,
815 F.3d 846, 879 (D.C. Cir. 2016) (per curiam); United States
v. Maxwell, 643 F.3d 1096, 1100 (8th Cir. 2011); United States
v. Parker, 553 F.3d 1309, 1314-15 (10th Cir. 2009); see also
United States v. Gibbs, 190 F.3d 188, 218 (3d Cir. 1999) (hold-
ing to the same effect on plain-error review).
On this standard, the District Court here did not abuse
its discretion. As we detail more fully below, both RICO and

22
Hernandez, Kelly, Sistrunk, and Eatmon all argue the point
in some form. Cruz, Villega, and Atkinson invoke Rule 28(i).

48
drug-trafficking conspiracy are ultimately grounded in the gen-
eral principles of conspiracy law. The Latham evidence impli-
cates several of the Defendants and goes to their willingness to
engage in concerted illegal action, amounting at its most seri-
ous to murder. The argument that the evidence has nothing to
do with drug trafficking and the South Side-Parkway rivalry is
therefore inapposite. Conspiracy is a single crime, even if it
embraces a multitude of ends to be achieved over a period of
time, by means that are not themselves the subject of agree-
ment among the conspirators. See Frohwerk v. United States,
249 U.S. 204, 209-10 (1919); infra, Section V.B.1. In this light,
a reasonable person could agree with the District Court that the
Latham evidence serves directly to prove the existence of
RICO conspiracy among the Defendants.
The Defendants’ Rule 403 challenges also fail. The fact
that the evidence is intrinsic establishes its probative nature,
and as the District Court pointed out, any evaluation of preju-
dicial effect here must be considered in the context of the to-
tality of the evidence produced. “The jury,” the District Court
observed, “has heard extensive evidence of Defendants’ and
their alleged co-conspirators’ drug trafficking and gun posses-
sion, gang membership, multiple shootings directed at their ri-
vals, shootouts on public streets involving feuding rivals in
which children are shot and even killed, and evidence of mul-
tiple murders.” App. 15. We agree with this assessment, and
conclude that the District Court did not abuse its discretion in
balancing the probative value and danger of prejudice as it did.
2. Expert Testimony
It is well established that a district judge has a “general
‘gatekeeping’ obligation” with respect to all testimony based
on specialized knowledge of some form. Kumho Tire Co. v.
Carmichael, 526 U.S. 137, 141 (1999). Under Federal Rule of

49
Evidence 702, she must ensure that such testimony is both re-
liable and relevant, including under the standard laid down in
Rule 403. Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579,
594-95 (1993). The judge must also ensure that “an expert wit-
ness [does] not state an opinion about whether [a] defendant
did or did not have a mental state or condition that constitutes
an element of the crime charged or of a defense.” Fed. R. Evid.
704(b). The Defendants here challenge two of the District
Court’s decisions under these rules. We review those decisions
for abuse of discretion. United States v. Davis, 726 F.3d 434,
446 (3d Cir. 2013).
Sistrunk’s Tattoo
The second superseding indictment included allegations
that several South Side members were affiliated with the
Bloods. Prior to trial, the Government announced its intention
to have John Havens, a Special Agent with the Federal Bureau
of Investigation, testify as an expert on the Bloods, detailing
among other things their organization and symbols. Anticipat-
ing a challenge to this proffer, the District Court held a Daubert
hearing. And when during trial the motion to exclude came, the
District Court ruled in a memorandum opinion that most of it
was admissible, but it excluded (among other things) testimony
“as to any individual defendant except in the abstract.” D. Ct.
Dkt. No. 860, at 11.
In support of its Blood-affiliation allegations, the Gov-
ernment sought to introduce depictions of a tattoo on Sistrunk’s
left bicep that read: “Live By The 5, Die By The [symbol of a
gun].” App. 5127; Sistrunk App. 78. Special Agent Havens
would not be shown the tattoo, the Government assured, but he
would describe the significance of certain symbols, such as the
number five. Sistrunk’s attorney objected under Rule 403, ar-
guing that this singled out his client in contradiction of the

50
Daubert decision. The District Court admitted the evidence,
and Sistrunk now appeals.
We find no abuse of discretion in the District Court’s
decision. Cooperating witnesses identified Sistrunk as a Blood.
Further, according to testimony of Special Agent Endy, when
federal agents executed the search warrant of Sistrunk’s home,
they found a letter signed, “Hat Boy, Low Ridah, Brim,
Kanye.” App. 5016. Special Agent Endy testified that “Kanye”
was Sistrunk’s alias and that “Brim” was “a Blood set refer-
ence”—that is, a reference to a particular subgroup of Bloods.
App. 5016. Sistrunk’s argument that this testimony and evi-
dence was minimal when compared to the voluminous trial rec-
ord is irrelevant. At the very least, the testimony represents in-
dependent support, apart from the tattoo and Special Agent Ha-
vens’s testimony, for the Government’s theory was Sistrunk
was affiliated with the Bloods.
Nor did the District Court’s decision to admit the evi-
dence unfairly single out Sistrunk in contradiction of the Daub-
ert ruling. Under that decision, Special Agent Havens would
not have testified as to Sistrunk in particular; the tattoo would
have been introduced after Special Agent Havens’s testimony,
and the jury would have been allowed to infer, or not infer, a
connection between the tattoo and the significance of the num-
ber five among certain Bloods. In fact, the point arguably be-
came explicit only through the efforts of Sistrunk’s attorney,
who on cross-examination presented Special Agent Havens
with a picture of the tattoo. Given this course of events, we are
comfortable that a reasonable person could adopt the District
Court’s view.

51
 The De La Cruz Criteria
One of the defense’s principal expert witnesses was Dr.
Jesse De La Cruz, a former gang member who earned a doc-
toral degree studying the gangs of Stockton, California. While
conducting that research, he developed a set of eight character-
istics common to the gang members he studied. Upon comple-
tion of his degree, Dr. De La Cruz began to testify as an expert
witness, determining whether a criminal defendant possessed
all or most of the characteristics. He interviewed all twelve de-
fendants and was prepared to say whether they met his criteria.
The Government challenged that proposed testimony
under Rule 704(b). It argued that Dr. De La Cruz could discuss
the eight characteristics and other matters, but that application
of the characteristics to the defendants would “go directly to
the intent of a particular person to be a member of a gang.”
App. 5752. The District Court agreed. It ruled that Dr. De La
Cruz could provide an “overview of gang activities” as a re-
sponse to Special Agent Havens, but that he could not discuss
whether the defendants met the eight criteria. App. 5754. That,
the District Court said, would amount to “testi[mony] as to a
person’s mental state or condition,” and the danger for preju-
dice was substantial in comparison with its limited probative
value. App. 5754-55. Joined by five others,23 Atkinson con-
tends that the District Court erred in excluding the testimony.
This was not reversible error. It may be true that Dr. De
La Cruz’s application of the eight criteria would not have con-
stituted “the last step in the inferential process—a conclusion
as to the [defendants’] mental state.” United States v. Watson,
260 F.3d 301, 309 (3d Cir. 2001) (citation omitted). As we de-

23
Cruz, Hernandez, Villega, Sistrunk, and Eatmon.

52
scribe in Section V.B.1 below, a RICO enterprise may still ex-
ist even if it does not amount to a gang, nor does gang mem-
bership in itself prove RICO conspiracy. Yet that distinction
illustrates the problematic nature of the testimony. The proba-
tive value was minimal unless one associates gang membership
with RICO conspiracy, and so any testimony to that effect
would have served, as the District Court said, only to “confuse
and mislead the jury.” App. 5755. “The trial judge has broad
discretion to admit or exclude expert testimony, based upon
whether it is helpful to the trier of fact.” Gibbs, 190 F.3d at
211. In this light, we cannot say the District Court abused its
discretion in excluding the testimony.
V. SUFFICIENCY OF THE EVIDENCE
We turn now to a series of interlocking challenges to the
sufficiency of the evidence supporting the jury’s verdicts. The
operative indictment charged all the Defendants in Counts I, II,
and III: RICO conspiracy, 18 U.S.C. § 1962(d); drug-traffick-
ing conspiracy, 21 U.S.C. § 846; and drug trafficking, 21
U.S.C. § 841(a), respectively. Seven Defendants—Cruz, Her-
nandez, Villega, Kelly, Atkinson, Sistrunk, and Eatmon—were
convicted on Count I, and each now contests his verdict.24
These same seven, plus Rice and Schueg, were convicted on
Counts II and III.25 All nine had drug quantities of 5 kilograms
or more of powder cocaine and 280 grams or more of crack
cocaine attributed to them, thus raising their mandatory mini-
mum term of imprisonment to 10 years and the maximum term

24
Cruz, Hernandez, Sistrunk, and Eatmon all argue the issue.
Villega, Kelly, and Atkinson raise it through Rule 28(i).
25
Williams was also convicted on Count III. He appeals only
his sentence, on grounds other than drug quantity. See infra
Section VI.A.1.

53
to life. See 21 U.S.C. § 841(b)(1)(A). Six of these nine—Her-
nandez, Villega, Rice, Kelly, Sistrunk, and Eatmon—now
challenge the verdicts on Counts II and III.26
For the reasons that follow, we will affirm the judg-
ments of conviction. We also shall affirm the jury’s Count II
drug-quantity verdicts insofar as they bear on the Defendants’
statutory maximum terms of imprisonment.
A. The Rowe Error
We begin with the legal framework governing our in-
quiry. Nearly three and a half years after trial, and after all the
Defendants had been sentenced, our Court in United States v.
Rowe, 919 F.3d 752 (3d Cir. 2019), clarified the effect of Al-
leyne v. United States, 570 U.S. 99 (2013), upon the distribu-
tion and possession elements of § 841(a)(1). We held that the
provisions of § 841(b)(1)(A) and (b)(1)(B) attach to each dis-
crete act of distribution or possession because they specify
facts that increase the statutory penalty, and so, under Alleyne,
constitute an “element of a distinct and aggravated crime,” 570
U.S. at 116, that must be submitted to the jury, see Rowe, 919
F.3d at 759. As a result, the jury may not “combine the amounts
distributed or possessed” at discrete instances to find the drug

26
On Count II, Hernandez, Villega, Rice, and Sistrunk argue
the issue in some form, while Kelly and Eatmon raise it through
Rule 28(i). Hernandez, Villega, and Rice also argue Count III;
Kelly, Sistrunk, and Eatmon all invoke Rule 28(i). In an ad-
dendum to his opening brief, Hernandez challenged his con-
viction on Count VI by incorporating without explanation Vil-
lega’s argument as to Count II. This was an improper adoption.
At least in this context, we fail to see how a Rule 28(i) incor-
poration of a co-defendant’s argument on a different count is
applicable, absent elaboration that was not provided.

54
quantities specified in § 841(b)(1)(A) and (b)(1)(B). Id. at 761.
The parties agree that under Rowe the evidence was in-
sufficient to support the Count III verdicts attributing to the
Defendants the § 841(b)(1)(A) quantities. The jury here was
charged on an aggregation theory of § 841(a)(1). The parties
contest, however, our standard of review of that error. Further,
two Defendants argue that Rowe also affects the jury’s drug-
quantity attributions on Count II—drug-trafficking conspiracy.
We will address each argument in turn. We conclude that re-
medial action on the Count III error is warranted only if the
Defendants’ terms of imprisonment would have been different
absent the error. Further, we conclude that an aggregation error
did occur on Count II, but only as it regards the Defendants’
mandatory minimum terms of imprisonment, and that the same
standard of review applies as for the Rowe error on Count III.
1. Standard of Review
When a new rule is issued during the pendency of a di-
rect criminal appeal, it is the appellate court’s duty to “apply
the law in effect at the time it renders its decision.” United
States v. Johnson, 899 F.3d 191, 199 (3d Cir. 2018) (quoting
Henderson v. United States, 568 U.S. 266, 271 (2013)). But
that does not necessarily determine our standard of review. Sis-
trunk contends that his Rule 29 motion at the close of the Gov-
ernment’s case in chief sufficiently preserved the issue. We
disagree.
The standard for preserving an argument on a Rule 29
motion remains an open question in our circuit. In United
States v. Joseph, 730 F.3d 336 (3d Cir. 2013), we drew a dis-
tinction between “issues” and “arguments,” noting that the for-
mer “can encompass more than one of the latter.” Id. at 340.
We then held that, in the evidence-suppression context, “for
parties to preserve an argument for appeal, they must have

55
raised the same argument in the District Court—merely raising
an issue that encompasses the appellate argument” results in
waiver of the argument. Id. at 337 (emphases omitted). The
Government invites us to apply this standard here.
Nearly all of our sister circuits, though, have settled on
a somewhat different standard. One has said that when a de-
fendant makes “general motions pursuant to Rule 29 for ac-
quittal, generally arguing that the government presented insuf-
ficient evidence,” he has “preserved his sufficiency claims for
appeal.” United States v. Hoy, 137 F.3d 726, 729 (2d Cir.
1998). Others have maintained that “[w]hen a defendant raises
specific grounds in a Rule 29 motion, grounds that are not spe-
cifically raised” are subject to some form of plain-error review,
if not waived, on appeal. United States v. Chong Lam, 677 F.3d
190, 200 (4th Cir. 2012) (emphasis omitted).27 A plurality of
circuits has explicitly adopted both of these standards.28 Only
the Fifth Circuit applies a Joseph-like standard in the Rule 29
context. See United States v. McDowell, 498 F.3d 308, 312-13
(5th Cir. 2007).
We think uniformity in federal criminal practice has
value, and so we decline to import Joseph wholesale here. It is
unnecessary, though, to diverge too far from Joseph and hold

27
See also United States v. Samuels, 874 F.3d 1032, 1036 (8th
Cir. 2017); United States v. Baston, 818 F.3d 651, 663-64 (11th
Cir. 2016).
28
See United States v. Porter, 886 F.3d 562, 566 (6th Cir.
2018); United States v. Marston, 694 F.3d 131, 134 (1st Cir.
2012); United States v. Hosseini, 679 F.3d 544, 550 (7th Cir.
2012); United States v. Graf, 610 F.3d 1148, 1166 (9th Cir.
2010); United States v. Goode, 483 F.3d 676, 681 (10th Cir.
2007); United States v. Spinner, 152 F.3d 950, 955 (D.C. Cir.
1998).

56
that a broadly stated Rule 29 motion preserves all arguments
bearing on the sufficiency of the evidence. It is enough to ac-
cept here that when a Rule 29 motion raises specific grounds,
or arguments (in the Joseph sense), all such arguments not
raised are unpreserved on appeal. Sistrunk’s motion raised a
narrow factual argument regarding the testimony of a witness.
That is a specific ground distinct from the Rowe argument, ren-
dering the latter unpreserved. Our principal divergence from
Joseph comes in how to treat the error: we will review for plain
error.29
The parties agree that Olano’s first and second prongs
are satisfied, and so our focus is on the substantial-rights in-
quiry. In Vazquez, we confronted a § 841 violation of Apprendi
v. New Jersey, 530 U.S. 466 (2000): “the drug quantity [wa]s
not found by a jury beyond a reasonable doubt and the defend-
ant’s sentence under § 841 exceed[ed] 20 years.” 271 F.3d at
98. Because this violation involved both a sentencing error and
a trial error, our substantial-rights inquiry asked whether “the
sentence would have been the same absent the trial error.” Id.

29
The circuits are more divided on this question than on the
preservation standard itself. One accepts full waiver, Porter,
886 F.3d at 566; two review for “a manifest miscarriage of jus-
tice,” Chong Lam, 677 F.3d at 200 n.10; Graf, 610 F.3d at
1166; one looks for “clear and gross injustice,” Marston, 694
F.3d at 134; and five review for plain error, Samuels, 874 F.3d
at 1036; Baston, 818 F.3d at 664; Hosseini, 679 F.3d at 550;
Goode, 483 F.3d at 681; Spinner, 152 F.3d at 955. Our Court
has in the past reviewed unpreserved sufficiency arguments for
plain error. See United States v. Husmann, 765 F.3d 169, 172,
173 n.2 (3d Cir. 2014); United States v. Sussman, 709 F.3d 155,
162 (3d Cir. 2013). Given this practice, and the nature of the
error here, we think plain-error review is appropriate.

57
at 101 (emphases omitted).
A similar approach is appropriate here. A Rowe error’s
principal effect goes to the sentence imposed. The “aggravated
crime,” Alleyne, 570 U.S. at 116, charged in Count III encom-
passes the “lesser included offense” of a “[v]iolation of
§ 841(a)(1),” Burrage v. United States, 571 U.S. 204, 210 n.3
(2014). The default penalty for that offense is specified in
§ 841(b)(1)(C). As a result, any prejudice arising from the
Rowe error concerns the length of the Defendants’ incarcera-
tion rather than the integrity of the general verdicts against
them.30 And we may assume that any additional day an error
causes a person to spend in prison affects his substantial rights.
See, e.g., Molina-Martinez, 136 S. Ct. at 1345.
To determine whether the Defendants’ sentences would
have been different absent the Rowe error, we may look in the
first instance to the evidence supporting the verdicts on Count
II—drug-trafficking conspiracy under 21 U.S.C. § 846.31 As
noted, the six challengers to Count III are the same six who

30
No Defendant challenges his conviction of the lesser in-
cluded offense of simple distribution. The Rowe error therefore
did not affect the Defendants’ substantial rights regarding the
$100 assessment for felony convictions pursuant to 18 U.S.C.
§ 3013(a)(2)(A). See United States v. Tann, 577 F.3d 533, 539-
40 (3d Cir. 2009).
31
RICO caps violations at 20 years’ imprisonment unless “the
violation is based on a racketeering activity for which the max-
imum penalty includes life imprisonment.” 18 U.S.C.
§ 1963(a). Here, the alleged predicate offenses were violations
of § 841(a)(1) at the § 841(b)(1)(A) quantities—for which the
maximum penalty is life imprisonment. The conceded Rowe
error therefore necessarily infects the validity of the sentences
on Count I.

58
contest their convictions on Count II. These six were sentenced
to concurrent terms of imprisonment on both counts. See supra
Section I.D. If the evidence is sufficient to support the jury’s
drug-quantity attributions on Count II—and, in particular, the
resulting maximum term of imprisonment under
§ 841(b)(1)(A)32—then vacating the drug-quantity verdicts on
Count III would not result in reduced sentences. It would,
therefore, be unnecessary for us to correct the Rowe error.
2. Section 846 Conspiracy and Drug Quantity: The Legal
Standard
Hernandez and Sistrunk contend that Rowe and Alleyne
also affect our evaluation of the evidence supporting the drug-
quantity verdicts on Count II. In particular, they argue that
those decisions either transformed drug quantity into a mens
rea element of § 846, or barred the aggregation of drug quan-
tity for sentencing purposes under § 846. We reject the first
argument, but qualifiedly agree on the second. We hold that a
jury, in determining drug quantity for purposes of the manda-
tory minimum term of imprisonment, may attribute to a de-
fendant only those quantities involved in violations of § 841(a)
that were within the scope of the conspiracy, or in furtherance
of it, and were reasonably foreseeable to the defendant as a nat-
ural consequence of his unlawful agreement.
Mental Element
Section 846 does not demand that a person conspire to
distribute a particular quantity of a controlled substance. To see
why, we must begin with the underlying statute. Under

32
The statutory maximum term under § 841(b)(1)(C) is still
greater than § 841(b)(1)(A)’s mandatory minimum, absent
other aggravating facts—such as a prior serious drug felony
conviction—that would apply anyway under (b)(1)(C).

59
§ 841(a)(1), “it shall be unlawful for any person knowingly or
intentionally . . . to . . . distribute, or . . . possess with intent to
. . . distribute, . . . a controlled substance.” This is the core of-
fense—the interdiction backed by the state’s claim to a monop-
oly of legitimate physical violence. Section 841(b) makes this
clear: it describes the penalties to be imposed upon “any person
who violates subsection (a) of this section.” 21 U.S.C.
§ 841(b). Properly speaking, then, a person who engages in
drug trafficking violates § 841(a), and the penalty for that vio-
lation is to be determined according to § 841(b), which pro-
vides both a default penalty and heightened penalties based on
certain additional factual findings. As a result, it is unnecessary
for the jury to find that the defendant knew the quantity of the
controlled substance he was distributing, or possessing with in-
tent to distribute, at a given time. It is enough that the knowing
or intentional distribution or possession occurred; the quantity
is a factual finding that goes to the sentence to be imposed. See
Burrage, 571 U.S. at 210-11 (interpreting § 841(b)(1)(C)’s “re-
sults from” enhancement as “impos[ing] . . . a requirement of
actual causality,” rather than legal causality, and thus as requir-
ing a factual finding of but-for causation); United States v.
Dado, 759 F.3d 550, 570 (6th Cir. 2014).
This interpretation is consistent with Apprendi and Al-
leyne. The Court in those cases operated on an expansive defi-
nition of “crime” according to its “invariable linkage” with
punishment, Apprendi, 530 U.S. at 478, rather than specifically
the conduct and mental state deemed illegal. Yet the decisions
did not fundamentally affect legislative authority to define a
crime’s elements. In Apprendi, for example, the Court noted
that traditionally, an indictment under a criminal statute that
“annexe[d] a higher degree of punishment to a common-law
felony, if committed under particular circumstances,” needed
to charge both “the circumstances of the crime and the intent

60
of the defendant at the time of commission,” and “the circum-
stances mandating [the higher] punishment.” Id. at 480 (quot-
ing John Archbold, Pleading and Evidence in Criminal Cases
51 (15th ed. 1862)). Both were “essential elements to be al-
leged,” id., but a prosecutor could fail to prove the latter and
still prove that the felony had been committed, id. at 480-81
(citing Archbold, supra, at 188). As a result, although bundled
in the broader concept of an “aggravated” crime, the statutory
definitions of “[t]he core crime” and the “triggering” fact re-
main the same. Alleyne, 570 U.S. at 113. In the context of
§ 841(a) and (b), that means the defendant need not con-
sciously cognize the amount he is distributing in order to vio-
late the law.
The same logic applies to drug-trafficking conspiracies
under § 846. The statute provides: “Any person who . . . con-
spires to commit any offense defined in this subchapter shall
be subject to the same penalties as those prescribed for the of-
fense, the commission of which was the object of the . . . con-
spiracy.” In the case of a drug-trafficking conspiracy, the “of-
fense” conspired is a violation of § 841(a), and the penalty for
this distinct crime—conspiracy to violate § 841(a)—is pro-
vided in § 841(b). For the same reason, then, that drug quantity
is not a mens rea element under § 841(a), it is not one under
§ 846.
Drug-Quantity Aggregation
The Defendants alternatively argue that just as Rowe
and Alleyne bar the aggregation of drug quantity for discrete
violations of § 841(a)(1), so they also bar aggregation for vio-
lations of § 846. The Government responds by referring to
United States v. Gori, 324 F.3d 234 (3d Cir. 2003), for the
proposition that the penalty for drug-trafficking conspiracy un-
der § 846 can be calculated according to the total amount of

61
drugs in the conspiracy. We agree with neither side fully. When
determining drug quantity for purposes of a defendant’s man-
datory minimum sentence, a jury must follow the ordinary lim-
itations on co-conspirator liability. Because that principle was
not followed here, we conclude that an error occurred on the
Count II drug-quantity verdicts.
In Gori, we recognized that the general principles of
conspiracy law may influence a defendant’s sentencing expo-
sure under § 846. When Congress borrows a legal term of art
in a criminal law, it is presumed to “know[] and adopt[] the
cluster of ideas that were attached” to that term and “the mean-
ing [the term’s] use will convey to the judicial mind,” absent
provision to the contrary. Morissette v. United States, 342 U.S.
246, 263 (1952). Section 846 is a law of this type, and so our
interpretation of it ought, where relevant, to have reference to
the “well-established principles,” Salinas v. United States, 522
U.S. 52, 63 (1997), of conspiracy law. See, e.g., Smith v. United
States, 568 U.S. 106, 110-11 (2013).
It is elementary that the “agreement to commit an of-
fense does not become several conspiracies because it contin-
ues over a period of time.” Braverman v. United States, 317
U.S. 49, 52 (1942). “[A] single continuing agreement to com-
mit several offenses” is equally a violation of the relevant con-
spiracy statute as a one-off agreement to commit a single of-
fense. Id.; see also United States v. Kissel, 218 U.S. 601, 607
(1910). Gori simply applied this principle in the context of a
§ 846 drug-trafficking conspiracy: one can conspire to violate
§ 841(a) multiple times, and this may constitute a single viola-
tion of § 846. 324 F.3d at 237. Moreover, because § 846 ties
its penalty to that of the substantive offense, and because, by
our foregoing logic, it is § 841(a) specifically that is conspired
to be violated, Gori’s interpretation of how to penalize a multi-
offense drug-trafficking conspiracy remains good law.

62
 Yet, importantly, Gori concerned the aggregation of
drug quantities arising from the offenses of the same defend-
ant. See 324 F.3d at 236. Equally central to conspiracy law is
the concept of co-conspirator liability. “It has always been, . .
. and is still, the law that, after prima facie evidence of an un-
lawful combination has been introduced, the act of any one of
the co-conspirators in furtherance of such combination may be
properly given in evidence against all.” Bannon v. United
States, 156 U.S. 464, 469 (1895). The “unlawful agreement
contemplated precisely what was [to be] done,” it “was formed
for the purpose” of committing a crime or crimes, and so the
“act of one partner in crime is attributable to all.” Pinkerton v.
United States, 328 U.S. 640, 647 (1946). Although thus ex-
panding liability, this logic contains its own limiting principle:
the act must be “done in furtherance of the conspiracy,” or “fall
within the scope of the unlawful project.” Id. at 647-48. A
“ramification[] of the plan which could not be reasonably fore-
seen as a necessary or natural consequence of the unlawful
agreement” does not bind the co-conspirator. Id. at 648. “No-
body is liable in conspiracy except for the fair import of the
concerted purpose or agreement as he understands it.” United
States v. Peoni, 100 F.2d 401, 403 (2d Cir. 1938) (L. Hand, J.).
These principles inform the extent of a defendant’s sen-
tencing exposure under § 846. In a post-Apprendi case, we held
that in prosecutions of multi-person drug-trafficking conspira-
cies, “[t]he [jury’s] finding of drug quantity for purposes of de-
termining the statutory maximum is . . . to be an offense-spe-
cific, not a defendant-specific, determination.” United States v.
Phillips, 349 F.3d 138, 143 (3d Cir. 2003), vacated on other
grounds sub nom. Barbour v. United States, 543 U.S. 1102
(2005). In other words, the jury finds only the quantity attribut-
able to “the conspiracy as a whole,” and then the sentencing

63
judge determines “the drug quantity attributable to each de-
fendant and sentence[s] him or her accordingly, provided that
the sentence does not exceed the applicable statutory maxi-
mum.” Id. “Accomplice attribution,” we recognized long be-
fore Phillips, “often results in a dramatic increase in the
amount of drugs for which the defendant is held accountable,
which translates directly into a dramatic increase in the sen-
tence.” United States v. Collado, 975 F.2d 985, 995 (3d Cir.
1992). And so, “at sentencing, it is essential for courts to con-
duct ‘a searching and individualized inquiry into the circum-
stances surrounding each defendant’s involvement in a con-
spiracy to ensure that the defendant’s sentence accurately re-
flects his or her role.’” United States v. Metro, 882 F.3d 431,
439 (3d Cir. 2018) (alterations omitted) (quoting Collado, 975
F.2d at 995).
Phillips’s holding did not apply to mandatory minimum
sentences. We adopted in that case the reasoning of three of
our sister circuits, see Phillips, 349 F.3d at 141-42 (citing
United States v. Knight, 342 F.3d 697, 710-11 (7th Cir. 2003);
United States v. Turner, 319 F.3d 716, 722-23 (5th Cir. 2003);
and Derman v. United States, 298 F.3d 34, 42-43 (1st Cir.
2002)), and those courts do not employ a conspiracy-wide ap-
proach in the context of mandatory minimums, see United
States v. Haines, 803 F.3d 713, 741-42 & n.9 (5th Cir. 2015);
United States v. Colón-Solís, 354 F.3d 101, 103 (1st Cir. 2004);
Knight, 342 F.3d at 711. Phillips said nothing to the contrary,
consistent with Collado: the jury sets the maximum according
to the total amount of drugs in the conspiracy, and the sentenc-
ing judge conducts an individualized inquiry to determine the
penalty for each co-conspirator.
Alleyne alters this regime. Since that decision, several
circuits—including the First and the Fifth—have held that the

64
jury, in determining (as Alleyne requires) drug quantity for pur-
poses of the mandatory minimum, may attribute to a defendant
only that “quantity which was within the scope of the agree-
ment and reasonably foreseeable to him.” United States v.
Dewberry, 790 F.3d 1022, 1030 (10th Cir. 2015) (internal quo-
tation marks omitted); see also United States v. Stoddard, 892
F.3d 1203, 1221 (D.C. Cir. 2018); Haines, 803 F.3d at 740;
United States v. Rangel, 781 F.3d 736, 742-43 (4th Cir. 2015);
United States v. Pizarro, 772 F.3d 284, 292-93 (1st Cir.
2014).33
We adopt here a similar, though not the same, approach.
The jury, when determining drug quantity for purposes of the
mandatory minimum, may attribute to a defendant only those
quantities involved in violations of § 841(a) that were within
the scope of, or in furtherance of, the conspiracy and were rea-
sonably foreseeable to the defendant as a consequence of the
unlawful agreement.34 We take this approach for two reasons.
First, it follows from the basic principles of our prece-

33
The Sixth Circuit has adopted the conspiracy-wide approach
for statutory minimum and maximum sentences. See United
States v. Gibson, No. 15-6122, 2016 WL 6839156, at *1 (6th
Cir. Nov. 21, 2016) (citing United States v. Robinson, 547 F.3d
632 (6th Cir. 2008)), aff’d by an equally divided court, 874
F.3d 544 (6th Cir. 2017) (mem).
34
The quantity of drugs for which conspirators can be held ac-
countable is not limited to amounts distributed or possessed
with intent to distribute. It also includes amounts that conspira-
tors agreed to distribute or possess with intent to distribute,
even if those amounts were not actually distributed or pos-
sessed.

65
dent. In Rowe, we acknowledged that because the drug quanti-
ties specified in § 841(b)(1)(A) and (b)(1)(B) increase the man-
datory minimum, they constitute facts that must be submitted
to the jury for each violation of § 841(a)(1). Gori is consistent
with Rowe because conspiracy law encompasses a continuing
agreement to commit several offenses, and so the penalty for a
violation of § 846 is appropriately calculated according to the
aggregate drug quantity involved in a defendant’s continuous
execution of the unlawful agreement. Under Alleyne, the jury
must determine this quantity to set the mandatory minimum.
Our holding here follows from the same rationale, applying to
this landscape another dimension of conspiracy law—co-con-
spirator liability—that must be considered by the jury. Where
Gori held that the drug quantities involved in a single conspira-
tor’s multiple violations of § 841(a) may be aggregated for pur-
poses of his sentence, we hold that the quantities involved in
the § 841(a) violations of multiple conspirators may be aggre-
gated for determining the mandatory minimum of any one con-
spirator, subject to the ordinary limitations on co-conspirator
liability.35

35
Pinkerton concerned liability for a distinct substantive of-
fense committed by a co-conspirator in furtherance of the con-
spiracy, rather than liability for the conspiracy offense itself.
However, its holding was simply an extension of an already
well-established principle that the act of a co-conspirator in
furtherance of the scheme is the act of all for purposes of con-
spiracy liability. See Pinkerton, 328 U.S. at 647. Our holding
here applies that idea to the § 846 drug-trafficking context. Fur-
ther, we think Pinkerton’s limitations on co-conspirator liabil-
ity apply to liability not only for a co-conspirator’s substantive
offense, but also under the relevant conspiracy statute. See,
e.g., Peoni, 100 F.2d at 403.

66
 Second, the approach is most consistent with our pre-
Alleyne regime. Phillips ensured that the jury would set the
maximum term a defendant could spend in prison, leaving it to
the judge to determine each co-conspirator’s individual sen-
tencing exposure under § 841(b). Here we transfer some of that
latter inquiry to the jury, as Alleyne requires. Yet in doing so,
we must necessarily alter it. Under Collado, the judge at sen-
tencing must “consider whether the amounts distributed by the
defendant’s co-conspirators . . . were reasonably foreseeable in
connection with the criminal activity the defendant agreed to
undertake.” 975 F.2d at 995 (internal quotation marks omitted)
(emphasis added). But as we have said, drug quantity is not a
mens rea element under § 846, and co-conspirator liability ex-
tends to acts or omissions that are reasonably foreseeable as a
consequence of the unlawful agreement. Accordingly, we think
the proper inquiry is to determine the violations of § 841(a)
within the scope of the conspiracy, or in furtherance of it, that
were reasonably foreseeable to the defendant as a natural result
of his unlawful agreement. All drug quantities involved therein
are attributable to the defendant.36

36
Collado’s specification that drug quantity itself needed to be
reasonably foreseeable was based on an application note of
U.S.S.G. § 1B1.3. And “[w]e have . . . explained that the con-
duct a defendant is typically held responsible for under the
guidelines is not coextensive with conspiracy law.” Metro, 882
F.3d at 439 (internal quotation marks omitted). Moreover, in
2015, the Sentencing Commission amended the relevant appli-
cation note so that it now reads: “With respect to offenses in-
volving contraband (including controlled substances), the de-
fendant is accountable [for] . . . all quantities of contraband that
were involved in transactions carried out by other participants,

67
 We thus agree with Hernandez and Sistrunk that an er-
ror occurred as to Count II. The jury rendered its verdicts by
considering only the amount of drugs involved in the conspir-
acy as a whole. But for the same reasons given above with re-
spect to the Rowe error on Count III—the drug-trafficking
count—this argument was not preserved in the Defendants’
Rule 29 motions, and so our review is for plain error. We may
assume that Olano’s second prong is satisfied. On the third
prong, our logic with respect to the Rowe error applies simi-
larly here. The error goes to the sentences imposed, and be-
cause (as we hold below) the Count II verdicts otherwise stand,
we may determine whether there is “a reasonable probability
that, but for the error claimed,” the Defendants’ terms of im-
prisonment would have been different. Dominguez Benitez,
542 U.S. at 82 (alteration and citation omitted).37 Further,
given our conclusions in Part VI below, with one exception,38
the Defendants’ sentences include incarceration in excess of
§ 841(b)(1)(A)’s mandatory minimum. The error, then, did not
affect their substantial rights.
B. Count I: RICO Conspiracy
Having clarified the legal framework of our inquiry, we
now turn to the sufficiency of the evidence on Counts I and
II—RICO conspiracy and drug-trafficking conspiracy. Both
offenses may arise from the same set of facts because they fol-
low from the general principles of conspiracy law. Here, the
operative indictment incorporated its allegations at Count I as

if those transactions . . . were reasonably foreseeable in con-
nection with that criminal activity.” U.S.S.G. § 1B1.3 cmt. n.3
(emphasis added).
37
Our discussion above of the $100 assessment for felony con-
victions, see supra note 30, thus also applies here.
38
Hernandez. See infra Section VI.A.2.

68
the basis for its charge at Count II. And, as we shall see, the
evidence supporting the Count I convictions overlaps with that
supporting the convictions on Count II.39 We hold that a ra-
tional juror could have concluded that each of the Defendants
convicted on Count I was guilty as charged.40
1. The Elements of the Offense
Conspiracy Generally
The fountainhead of any criminal conspiracy is the
agreement: when “two or more . . . confederate and combine
together, by concerted means, to do that which is unlawful or
criminal.” Callan v. Wilson, 127 U.S. 540, 555 (1888). Under
both the RICO- and the drug-trafficking-conspiracy statutes,
18 U.S.C. § 1962(d) and 21 U.S.C. § 846, “the Government
must prove beyond a reasonable doubt that two or more people
agreed to commit a crime covered by the specific conspiracy
statute (that a conspiracy existed) and that the defendant know-
ingly and willfully participated in the agreement (that he was a
member of the conspiracy).” Smith, 568 U.S. at 110. The stat-
utes are therefore “even more comprehensive than the general
conspiracy offense in [18 U.S.C.] § 371” because they do not
require an overt act. Salinas, 522 U.S. at 63; see also United
States v. Shabani, 513 U.S. 10, 17 (1994).
Further, the RICO or drug-trafficking conspiracy may
continue over time and embrace a multitude of objects. Smith,

39
Of the six Defendants raising a sufficiency challenge on
Count II, only Rice was not convicted on Count I. We address
the evidence supporting his conspiracy conviction in Section
V.C.2 below.
40
We consider here only the sufficiency of the evidence sup-
porting the jury’s general verdicts on Count I—commission of
the substantive offense. See supra note 31.

69
568 U.S. at 111. It may exist even if an individual conspirator
“does not agree to commit or facilitate each and every part of
the” contemplated crime or crimes. Salinas, 522 U.S. at 63.
Nor even must the conspiracy actually achieve any or all of its
criminal ends. United States v. Rabinowich, 238 U.S. 78, 86
(1915). It is enough that the conspirator “intend[s] to further an
endeavor which, if completed, would satisfy all of the elements
of a substantive criminal offense.” Salinas, 522 U.S. at 65.
Thus involved, each conspirator is subject to the ordi-
nary principles of co-conspirator liability. Smith, 568 U.S. at
111 (citing Pinkerton, 328 U.S. at 646). And he continues to
be liable “up to the time of abandonment or success.” Kissel,
218 U.S. at 608. Indeed, “a defendant’s membership in the con-
spiracy, and his responsibility for its acts, endures even if he is
entirely inactive after joining it.” Smith, 568 U.S. at 114; see
also Callanan v. United States, 364 U.S. 587, 593 (1961)
(“Group association for criminal purposes often, if not nor-
mally, makes possible the attainment of ends more complex
than those which one criminal could accomplish.”). Once the
prosecution has proved both the existence of a conspiracy
across a period of time and the defendant’s participation in that
conspiracy, the burden falls on the defendant to establish his
withdrawal prior to the completion of the period. Smith, 568
U.S. at 113. If he does not show “some [affirmative] act to dis-
avow or defeat the purpose” of the conspiracy, then he must
“incur the guilt” attendant upon its continuance. Hyde v. United
States, 225 U.S. 347, 369-70 (1912).
Section 1962(c)
Seven Defendants were convicted of conspiracy to vio-
late 18 U.S.C. § 1962(c). That provision declares in relevant
part:
It shall be unlawful for any person . . . associated

70
 with any enterprise engaged in, or the activities
of which affect, interstate . . . commerce, to con-
duct or participate, directly or indirectly, in the
conduct of such enterprise’s affairs through a
pattern of racketeering activity . . . .
For our purposes here, the final two elements are the most sig-
nificant: participation in (1) the conduct of an enterprise (2)
through a pattern of racketeering activity.
RICO defines an “enterprise” to “include[] any individ-
ual, partnership, corporation, association, or other legal entity,
and any union or group of individuals associated in fact alt-
hough not a legal entity.” 18 U.S.C. § 1961(4). In the present
cases, the enterprise was said to be the “Southside Gang,”
which was “a group of individuals associated in fact.” App. 25.
The jury was charged and returned its verdicts on this theory.
Despite considerable dispute at trial and in the briefs before us,
the term “gang” has no talismanic significance in the RICO
context. An association-in-fact enterprise, the Supreme Court
has said, is “an entity, for present purposes a group of persons
associated together for a common purpose of engaging in a
course of conduct.” United States v. Turkette, 452 U.S. 576,
583 (1981). This definition entails “at least three structural fea-
tures: a purpose, relationships among those associated with the
enterprise, and longevity sufficient to permit these associates
to pursue the enterprise’s purpose.” Boyle v. United States, 556
U.S. 938, 946 (2009). Beyond this the proof need not go: “an
association-in-fact enterprise is simply a continuing unit that
functions with a common purpose.” Id. at 948.
Next, “racketeering activity” is said to “mean[]” certain
criminal acts defined by statute, including “any offense involv-
ing . . . the felonious manufacture, importation, receiving, con-
cealment, buying, selling, or otherwise dealing in a controlled

71
substance.” 18 U.S.C. § 1961(1)(D). A “pattern of racketeering
activity” in turn “requires at least two acts of [such] activity, .
. . the last of which occurred within ten years (excluding any
period of imprisonment) after the commission of a prior act of
racketeering activity.” Id. § 1961(5); see H.J. Inc. v. Nw. Bell
Tel. Co., 492 U.S. 229, 239 (1989) (“[A] . . . prosecutor must
show that the racketeering predicates are related, and that they
amount to or pose a threat of continued criminal activity.” (em-
phasis in original)). Although the evidence establishing an en-
terprise and a pattern of racketeering activity “may in particular
cases coalesce,” the two elements themselves remain “at all
times” distinct. Turkette, 452 U.S. at 583.
Section 1962(d)
As relevant here, to be liable for RICO conspiracy under
§ 1962(d), a defendant must “intend to further an endeavor
which, if completed, would satisfy all of the elements of
[§ 1962(c)].” Salinas, 522 U.S. at 65. That endeavor may be
both the enterprise and the conspiracy, for the two crimes can
be “coincident in their factual circumstances.” Id. It is a “per-
son,” not the enterprise itself, who violates § 1962(c) by “con-
duct[ing] or participat[ing]” in the enterprise’s affairs “through
a pattern of racketeering activity.” 18 U.S.C. § 1962(c); see
United States v. Bergrin, 650 F.3d 257, 267 (3d Cir. 2011) (cit-
ing H.J. Inc., 492 U.S. at 244). The nature of the liability there-
fore depends upon the circumstances. A defendant may be a
party to the enterprise, not violate § 1962(c), but still be liable
under § 1962(d). He need not “commit or agree to commit the
two or more predicate acts requisite to [§ 1962(c)].” Salinas,
522 U.S. at 65. Nor even, thanks to the absence of an overt-act
requirement, must one of his co-conspirators actually violate
§ 1962(c). See id. at 63. It is enough that the defendant “knew
about and agreed to facilitate the scheme” which at least would
have resulted in the satisfaction of § 1962(c)’s elements. Id. at

72
66; see also United States v. Fattah, 914 F.3d 112, 164 (3d Cir.
2019).
Thus, consistent with the general principles of conspir-
acy law recited above, conspiracy to violate § 1962(c) requires:
(1) that two or more persons agree to further an enterprise
whose activities affect or would affect interstate or foreign
commerce, and whose execution results or would result in a
person conducting or participating directly or indirectly in the
enterprise’s affairs through a pattern of racketeering activity;
(2) that the defendant was a party to or a member of this agree-
ment; and (3) that the defendant joined the agreement knowing
of its objectives and with the intention of furthering or facili-
tating them. See United States v. John-Baptiste, 747 F.3d 186,
207 (3d Cir. 2014).
2. The Evidence
In any review of the sufficiency of the evidence support-
ing a criminal conviction, “the relevant question is whether,
after viewing the evidence in the light most favorable to the
prosecution, any rational trier of fact could have found the es-
sential elements of the crime beyond a reasonable doubt.”
Jackson v. Virginia, 443 U.S 307, 319 (1979) (emphasis in
original). The Government may prove the existence of a con-
spiracy entirely through circumstantial evidence. United States
v. Kapp, 781 F.2d 1008, 1010 (3d Cir. 1986). In such instances,
we sustain the verdict if the proof “appears as a reasonable and
logical inference” from “evidence of related facts and circum-
stances.” United States v. Brodie, 403 F.3d 123, 134 (3d Cir.
2005) (citation omitted). And we “must credit all available in-
ferences in favor of the government.” Fattah, 914 F.3d at 162
(citation omitted).
The Defendants—Cruz, Hernandez, Villega, Kelly, At-
kinson, Sistrunk, and Eatmon—contend that the alleged South

73
Side gang did not amount to an enterprise for purposes of
RICO liability. They point to testimony that the South Side was
simply a neighborhood where the Defendants grew up or lived;
that the drug dealing that occurred there amounted at best to
parallel conduct by independent actors; and that any violent in-
cidents were the product of personal “beefs.”
It is undeniable that the drug dealers operating on the
South Side during the indictment period did not constitute a
gang on the order of the Bloods or Crips. Nor was this a traf-
ficking operation to rival the ’Ndrangheta. Yet that is not what
RICO requires. The evidence need only support the conclusion
that each of these seven Defendants at least agreed to further a
continuing unit that functioned with a common illegal purpose.
Testimony showed that as early as 2002, Cruz, Hernan-
dez, and Kelly supplied crack to Atkinson and Eatmon in the
area around Maple and Duke. App. 3543-47, 3633-34; see also
App. 1503-07. Hernandez and Kelly also, it was said, helped
to introduce guns to the South Side, at least partially in re-
sponse to fighting with Parkway. App. 3553. A few years later,
Sistrunk began selling drugs at Maple and Duke. App. 3559-
60. By that time, however, Cruz, Hernandez, and Kelly had
been incarcerated, and so Atkinson, Eatmon, and Sistrunk,
among others, began collectively to traffic in large quantities
of crack. App. 3570-75, 3830-31; see also App. 2110-11;
3138-39. Their profits were all earned separately, App. 3817-
18, but nevertheless the men sometimes shared scales and
bought or fronted drugs among each other, App. 3574.
This association persisted into the next decade. See, e.g.,
App. 2456-57. In June 2011, while in prison, Villega told War-
ren Pillgreen to “straighten out that package,” referring to a
drug debt Pillgreen owed to Hernandez. App. 3016. A few
months later, shortly before Pillgreen’s release from prison,

74
Cruz engaged him to “commit an act of violence” to settle the
debt. App. 3018. By 2012, Cruz and Hernandez were still sup-
plying substantial amounts of crack, and Kelly was present for
these transactions. App. 3644-48. In September, Cruz, Hernan-
dez, Kelly, Atkinson, and Eatmon were involved in a physical
altercation between South Side and Parkway at Rutter’s gas
station. App. 3649-63. The Latham homicide occurred just
over two months later—an event, we have seen, in which Her-
nandez, Cruz, Kelly, and perhaps Sistrunk played a part. App.
36370-71; 3859-61. Finally, in early 2014, Villega’s floormate
at a halfway house worked with him to sell heroin, and occa-
sionally observed him with other individuals coming to and
from the house’s basement. App. 4513-16. When police later
searched the house, they discovered approximately 13.5 grams
of heroin and 61 grams of crack in the basement, and a photo-
graph in Villega’s bedroom of himself, Cruz, and Hernandez.
App. 4561, 4567-68.
A rational juror could conclude from this evidence—
and, more generally, from the entire body of evidence—that
each of the seven challengers agreed to further an enterprise
whose predominant common purpose was “making money”
through the sale of controlled substances, but which also occa-
sionally embraced related ends, such as “protecting its own
members and criminal schemes.” See Bergrin, 650 F.3d at 269.
As noted, the conspiracy and the enterprise need not be distinct,
and a continuing unit for purposes of RICO may exist even if
a given Defendant was not always active. See Boyle, 556 U.S.
at 948 (“[N]othing in RICO exempts an enterprise whose asso-
ciates engage in spurts of activity punctuated by periods of qui-
escence.”); see also Smith, 568 U.S. at 114. Here, each of the
Defendants persisted in the group’s concerted illicit activities
over an extended period of time, operating within the larger, if
“relatively loose and informal,” Bergrin, 650 F.3d at 269,

75
structure of the South Side’s drug blocks. Based on this evi-
dence, we cannot say that no rational juror would find the De-
fendants guilty of RICO conspiracy under § 1962(d).
C. Count II: Drug-Trafficking Conspiracy
We proceed, finally, to the evidence supporting the con-
victions on Count II. Six Defendants—Hernandez, Villega,
Rice, Kelly, Sistrunk, and Eatmon—challenge the sufficiency
of this evidence. We hold that a rational juror could have found
each of the challengers guilty under § 846 and attributed to him
the § 841(b)(1)(A) quantities for purposes of his statutory max-
imum term of imprisonment.
1. The Elements of the Offense
We have already described some of the basic principles
governing a defendant’s liability under § 846 for participation
in a drug-trafficking conspiracy. See supra Sections V.A.2,
V.B.1. Our precedent and the foregoing discussion establish
three basic elements that the jury must find beyond a reasona-
ble doubt.
First, there must be a conspiracy—an agreement among
two or more persons to achieve by concerted means an illegal
goal. It has long been settled in our Court that to prove a drug-
trafficking conspiracy, “the government must establish a unity
of purpose between the alleged conspirators, an intent to
achieve a common goal, and an agreement to work together
toward that goal.” Gibbs, 190 F.3d at 197. A conspiracy under
§ 846 becomes a drug-trafficking conspiracy when that com-
mon goal is a violation or violations of § 841(a). But “[t]he
government need not prove that each defendant knew all of the
conspiracy’s details, goals, or other participants.” United
States v. Bailey, 840 F.3d 99, 108 (3d Cir. 2016) (internal quo-
tation marks omitted). The agreement “is the essence of the of-

76
fense,” and “the presence of certain facts often provides cir-
cumstantial evidence of the underlying agreement.” United
States v. Pressler, 256 F.3d 144, 147 (3d Cir. 2001).
Second, the defendant must have been a member of the
conspiracy. He must be shown to have intended to further a
scheme whose execution he knew would or did result in the
commission of each element of the substantive offense. Under
this latter “knowledge” requirement, the government must
prove “that the defendant had knowledge of the specific objec-
tive contemplated by the . . . conspiracy.” United States v. Car-
aballo-Rodriguez, 726 F.3d 418, 425 (3d Cir. 2013) (en banc).
In the present context, that means he must have known that the
conspiracy would or did result in the distribution of a con-
trolled substance.
Although the evidence establishing the existence of a
conspiracy may coincide with proof of participation in that
conspiracy, “certain types of circumstantial evidence become
substantially more probative if it can be established that a con-
spiracy existed and the only remaining question is whether the
defendant was a part of it.” Pressler, 256 F.3d at 151. “[A]
simple buyer-seller relationship,” however, “without any prior
or contemporaneous understanding beyond the sales agree-
ment itself, is insufficient to establish that the buyer was a
member of the seller’s conspiracy.” Gibbs, 190 F.3d at 197.
Rather, the “buyer” is liable under § 846 only if direct or cir-
cumstantial evidence shows that he knew “he was part of a
larger operation.” United States v. Price, 13 F.3d 711, 728 (3d
Cir. 1994); see also Gibbs, 190 F.3d at 199-200 (listing several
factors for making this determination).
Third, if the indictment charges drug quantities pursuant
to § 841(b)(1)(A) or (b)(1)(B), then the statutory maximum
term of imprisonment is to be determined according to the

77
amount of drugs involved in the conspiracy as a whole. Phil-
lips, 349 F.3d at 143. The mandatory minimum, however, may
be determined only according to the aggregate quantity of
drugs involved in those violations of § 841(a) that were within
the scope of the conspiracy, or in furtherance of it, and were
reasonably foreseeable to the defendant as a natural conse-
quence of his unlawful agreement. See supra Section V.A.2.
2. The Evidence
We proceed generally according to the sufficiency-of-
the-evidence standard recited above. In cases of drug-traffick-
ing conspiracy, “the verdict must be upheld as long as it does
not ‘fall below the threshold of bare rationality.’” Caraballo-
Rodriguez, 726 F.3d at 431 (quoting Coleman v. Johnson, 566
U.S. 650, 656 (2012)).
The challengers contest the jury’s verdicts on two
grounds. First, they contend there was no evidence of an agree-
ment either to form a conspiracy or to join one. Second, they
dispute the evidence as to drug quantity. We consider each ar-
gument in turn.
Agreement
Our foregoing discussion establishes the common foun-
dation of RICO and drug-trafficking conspiracy in the general
principles of conspiracy law. The two offenses may be coinci-
dent in their factual circumstances, especially where the pattern
of racketeering activity involves “the felonious manufacture, .
. . buying, selling, or otherwise dealing in a controlled sub-
stance.” 18 U.S.C. § 1961(1)(D). In the present cases, our eval-
uation of the evidence supporting the Count I convictions of
Hernandez, Villega, Kelly, Sistrunk, and Eatmon also applies
here with regard to the requisite conspiratorial agreement. See
United States v. Rodríguez-Torres, 939 F.3d 16, 30 (1st Cir.
2019) (resolving defendants’ sufficiency challenges to drug-

78
trafficking-conspiracy convictions on the basis of a preceding
resolution of their sufficiency challenges to their RICO-con-
spiracy convictions).
The only Defendant to challenge his Count II conviction
who was not convicted on Count I is Rice. He argues that there
is insufficient evidence showing that he ever joined the charged
conspiracy—that he was, at most, a street-level dealer who
abandoned that lifestyle upon his release from prison in 2013.
The evidence in the record belies this argument. For example,
Cordaress Rogers testified that he, Rice, Atkinson, Sistrunk,
and Eatmon were “at one time in life . . . like[] brothers” and
would hang out and sell drugs together every day around Ma-
ple and Duke. App. 3571. This went beyond friendship to mu-
tual facilitation of drug trafficking. They would gather at each
other’s houses to sell drugs; they would buy drugs from each
other or front them to each other when one ran out; they would
share scales for measuring the drugs. App. 3572-74. They sold
primarily to dealers, rather than users. App. 3830-31. Further,
upon his release from prison in late July 2008, Rice returned to
the South Side. At that point, Jerrod Brown identified Rice as
handling guns and seeking retribution for the shooting of Jah-
keem Abney outside the night club in mid-March 2009. App.
2113-14. Finally, Brown also testified that sometime after May
2013—which would have been shortly after Rice’s release
from prison—Rice supplied him with crack. App. 2163-64.
Based on this evidence, a reasonable juror could con-
clude that Rice was consciously and willingly a part of a larger
drug-trafficking operation and remained so even after periods
of imprisonment. See Gibbs, 190 F.3d at 200.
Drug Quantity
A rational juror also could conclude that Hernandez,

79
Villega, Rice, Kelly, Sistrunk, and Eatmon were each respon-
sible, on a conspiracy-wide basis, for 280 grams or more of
crack cocaine. Rogers’s testimony alone indicated that in the
early years just after 2002, he received 1 kilogram of crack
from each of Hernandez, Kelly, and Cruz. App. 3633-34. At
that time, he was close with Atkinson, Rice, and Eatmon, who
were receiving drugs from Hernandez and Kelly in similar
quantities. App. 3543-45. Rogers also estimated that in later
years, when he, Atkinson, Rice, Sistrunk, and Eatmon worked
closely together, he would bring back from New York 500-
1000 grams of crack “[e]very couple of days.” App. 3573. He
testified that in this time he distributed and saw his friends dis-
tribute “many kilos of crack.” App. 3575. Moreover, to the ex-
tent that any of the Defendants were incarcerated and could not
have been present for the movement of these quantities, their
renewed drug dealing upon release from prison confirms their
continuing liability for acts in furtherance of the conspiracy,
even apart from the absence of an affirmative act of with-
drawal. See Hyde, 225 U.S. at 369-70.
Finally, as noted above, Villega aided Hernandez in the
collection of a drug debt by warning Pillgreen to “straighten
out that package.” App. 3016. Marquis Williams testified that
Villega fronted him 6 grams of heroin in 2013. App. 2443-44,
2655. By early 2014, Villega was still dealing heroin, App.
4513-16, and police later recovered about 13.5 grams of heroin
and 61 grams of crack from the basement Villega was seen to
frequent with others. App. 4561. In just that timeframe, from
2011 to 2014, Rogers testified that he received 156 grams of
crack from Cruz and Hernandez, App. 3645-46, and Marquis
Williams said he sold 50-gram quantities of crack on “several”
occasions, App. 2442. Based on this evidence alone, an attrib-
ution to Villega of over 280 grams of crack on a conspiracy-
wide basis does not fall below the threshold of bare rationality.

80
 ***
There was sufficient evidence upon which a rational ju-
ror could have concluded that these six Defendants were guilty
under § 846 and were responsible for 280 grams or more of
crack. Because we reach this conclusion, we further conclude
that the Rowe error on Count III did not affect their substantial
rights. Their statutory maximum terms of imprisonment would
have been life even if the Rowe error had not occurred.
VI. SENTENCING
The final category of issues concerns the sentences im-
posed in the years following the trial. All the Defendants chal-
lenge various aspects of those judgments.41 For the reasons that
follow, we will affirm the judgments of sentence of Williams
and Rice. But we will vacate Hernandez’s judgment of sen-
tence in full, the other Defendants’ judgments of sentence in
part, and remand the cases of Hernandez and Schueg for resen-
tencing proceedings consistent with this opinion.
A. Individual Challenges
1. Williams
Jabree Williams’s Presentence Report (PSR) recom-
mended a Guidelines range of 78-97 months in prison. The
District Court sentenced him to 60 months, the mandatory min-

41
Some Defendants have sought, pursuant to Rule 28(i), to
adopt sentencing challenges of others. However, general adop-
tions or ones that concern an argument specific to the arguing
party will not be regarded, if they are not accompanied by fur-
ther elaboration. We refuse to speculate on how an issue ap-
plies to a Defendant’s sentencing judgment when he himself
has declined to do so.

81
imum, based upon time served for two prior state drug convic-
tions. The Court also recommended that the Bureau of Prisons
credit Williams with an additional 13 months for time served
on a prior juvenile offense, and with approximately 28-29
months for time in federal custody.
Williams raises only one issue on appeal. The District
Court, he contends, should have credited the 13-month term
because 18 U.S.C. § 3584, as applied here, violates his Fifth
Amendment due process right. That provision states, in rele-
vant part: “if a term of imprisonment is imposed on a defendant
who is already subject to an undischarged term of imprison-
ment, the terms may run concurrently or consecutively.” 18
U.S.C. § 3584(a). Williams argues that the statute draws an ar-
bitrary distinction between discharged and undischarged sen-
tences. The Government counters that Williams did not raise
this issue contemporaneously, and that a reversible plain error
did not occur. Williams offers no reply, and there is no evi-
dence suggesting preservation. Our review, then, is for plain
error.
We need not address the merits of Williams’s constitu-
tional challenge to § 3584. For even if there was an error, it
was not plain. A “court of appeals cannot correct an error pur-
suant to Rule 52(b) unless the error is clear under current law.”
Olano, 507 U.S. at 734. Every court of appeals to have consid-
ered the issue, or a related challenge to U.S.S.G. § 5G1.3(b),
has rejected Williams’s argument. See United States v. Lucas,
745 F.3d 626, 630-31 (2d Cir. 2014) (per curiam); United
States v. Dunham, 295 F.3d 605, 610-11 (6th Cir. 2002);
United States v. Otto, 176 F.3d 416, 418 (8th Cir. 1999). Only
a district court, in another circuit, has held to the contrary.
United States v. Hill, 187 F. Supp. 3d 959, 965 (N.D. Ill. 2016).
Given the balance of such authority, it cannot be said the as-
sumed error here is “‘obvious’ or ‘clear under current law.’”

82
Vazquez, 271 F.3d at 100 (quoting Olano, 507 U.S. at 734). We
reserve for another day our own views on the merits.
2. Hernandez
At Hernandez’s sentencing hearing, his attorney, Mor-
ris, stated that “Mr. Hernandez does not desire to address the
court this morning. However, he did want me to say that he
wanted to thank his family for their support of him throughout
this process, and so we’d have nothing further beyond that.”
App. 289. The District Court accepted this submission, and,
after allowing the Government an opportunity to speak, an-
nounced its judgment. It did not address Hernandez personally,
and neither Morris nor the Government raised Hernandez’s
right to allocution. See Fed. R. Crim. P. 32(i)(4)(a)(ii); Green
v. United States, 365 U.S. 301, 305 (1961) (plurality opinion);
id. at 307 (Black, J., dissenting).
Hernandez now argues that he is entitled to resentencing
proceedings under United States v. Adams, 252 F.3d 276 (3d
Cir. 2001). The Government concedes the point, but it asserts
without elaboration that resentencing “should be limited to
providing Hernandez the opportunity to allocute should he so
desire.” Gov’t Br. at 212. We disagree. In Hill v. United States,
368 U.S. 424 (1962), the Supreme Court cited Van Hook v.
United States, 365 U.S. 609 (1961) (per curiam), for the appro-
priate remedy in direct appeals. 368 U.S. at 429 n.6. Van Hook
is a one-sentence opinion, stating: “The judgment is reversed
and the case remanded for resentencing in compliance with”
Rule 32 and Green. 365 U.S. at 609. This language provides
no indication of a limited remand, and our post-Adams cases
have not applied such a remedy. See United States v. Chapman,
915 F.3d 139, 147 (3d Cir. 2019); United States v. Paladino,
769 F.3d 197, 204 (3d Cir. 2014); United States v. Plotts, 359

83
F.3d 247, 251 (3d Cir. 2004). Hernandez is entitled to a resen-
tencing proceeding, with all its attendant considerations.42 See,
e.g., Pepper v. United States, 562 U.S. 476 (2011). However,
the District Court may, in its discretion, allow the Government
to offer new evidence. United States v. Dickler, 64 F.3d 818,
831-32 (3d Cir. 1995).
3. Kelly
Kelly brings several challenges to his concurrent life
sentences. Five of those challenges are unique to him. He as-
serts four procedural defects in the District Court’s decision,
and he claims that the sentences were substantively unreason-
able. We review procedural-soundness and substantive-unrea-
sonableness challenges for abuse of discretion. United States
v. Tomko, 562 F.3d 558, 567 (3d Cir. 2009) (en banc). Further,
“[w]e exercise plenary review of a district court’s interpreta-
tion of the Sentencing Guidelines and review its factual find-
ings for clear error.” United States v. Welshans, 892 F.3d 566,
573 (3d Cir. 2018). Four of the issues are meritless. The other
leaves Kelly’s sentence unaffected.
1. Dangerous-weapon enhancement. Kelly asserts that the
District Court erred in applying the two-level enhancement for
possession of “a dangerous weapon” in connection with a con-
trolled-substances offense. U.S.S.G. § 2D1.1(b)(1). We disa-
gree. The government can show possession simply “by estab-
lishing that a temporal and spatial relation existed between the
weapon, the drug trafficking activity, and the defendant.”
United States v. Napolitan, 762 F.3d 297, 309 (3d Cir. 2014)
(internal quotation marks omitted). If it does so, the burden

42
Because we reach this conclusion, we address neither Her-
nandez’s other sentencing challenges nor the effect of the man-
datory minimum error at Count II.

84
shifts to the defendant to show that “it is clearly improbable
that the weapon was connected with the offense.” U.S.S.G. §
2D1.1 cmt. n.11. Here, Cordaress Rogers testified that Kelly
supplied drugs to numerous younger dealers and helped to in-
troduce guns to the South Side, that a lot of people had guns,
and that guns were stashed on the blocks. The prevalence of
firearms was also described in other testimony. This evidence
establishes the requisite nexus, and Kelly gives no indication
of clear improbability.
2. Organizer or leader increase. Kelly contends that the
District Court erred in applying a four-level increase for being
“an organizer or leader of a criminal activity that involved five
or more participants.” U.S.S.G. § 3B1.1(a). When determining
whether to apply this enhancement, a court should consider,
among other things, “the nature of participation in the commis-
sion of the offense, . . . the degree of participation in planning
or organizing the offense, [and] the nature and scope of the il-
legal activity.” U.S.S.G. § 3B1.1 cmt. n.4. As just noted, the
evidence indicated that Kelly supplied a substantial amount of
crack to the younger generation of street-level dealers, associ-
ated with other key suppliers such as Cruz and Hernandez, and
helped to introduce guns into the South Side-Parkway rivalry.
In this light, we cannot say the District Court clearly erred in
imposing the enhancement.
3. Calculation of criminal-history score. Kelly next con-
tests his classification as a career offender for purposes of his
criminal-history category. Under the Guidelines, a defendant is
a career offender if he “has at least two prior felony convictions
of either a crime of violence or a controlled substance offense.”
U.S.S.G. § 4B1.1(a). An offense committed before the age of
18 may qualify “if it is classified as an adult conviction under
the laws of the jurisdiction in which the defendant was con-
victed.” U.S.S.G. § 4B1.2 cmt. n.1. Kelly argues that one of his

85
predicate convictions, a November 1994 conviction in New
York state court for attempted murder, was not so classified.
The District Court found that it was, based largely on a “Sen-
tence & Commitment” form of the New York Supreme Court,
Bronx County.
This finding was not clearly erroneous. As the District
Court pointed out, on the form there were two options after the
line “The defendant having been.” Gov’t Supp. App. 165. One
was “convicted of the crime(s) of”; the other, “adjudicated a
Youthful Offender.” The former was checked, suggesting
Kelly’s conviction was the same as that for an adult. At the
bottom of the form was written “YO denied.” The District
Court reasonably inferred that this meant “youthful offender
denied.” Kelly App. 518. Finally, simply because Kelly was
marked a “Juvenile Offender” on the form is not, under appli-
cable New York law, indicative of a non-adult conviction. See
In re Raymond G., 715 N.E.2d 486, 488 (N.Y. 1999); Matter
of Vega, 393 N.E.2d 450, 452-53 (N.Y. 1979).
4. Use-of-violence enhancement. Kelly points out that the
District Court failed to consider his objection to the two-level
enhancement under U.S.S.G. § 2D1.1(b)(2). The Government
essentially concedes the point, arguing only that the District
Court addressed Kelly’s use of violence when it rendered its
decision. But, of course, the sentencing judge must make “an
individualized assessment based on the facts presented.” Gall
v. United States, 552 U.S. 38, 50 (2007). Nevertheless, because
we reject Kelly’s other procedural challenges, this error does
not affect his total offense level.
5. Substantive reasonableness. “[I]f the district court’s
sentence is procedurally sound, we will affirm it unless no rea-
sonable sentencing court would have imposed the same sen-
tence on that particular defendant for the reasons the district

86
court provided.” Tomko, 562 F.3d at 568. In rendering its judg-
ment, the District Court said: “[Kelly’s] not here for an isolated
event, he’s here for a decade-long conspiracy that involved
multiple episodes of violence and harm to innocen[ts] in the
community . . . . The defendant was at the core of this enter-
prise and these violent acts.” Kelly App. 528. The District
Court noted Kelly’s “involve[ment] in drug and gang activity
from a very young age.” Kelly App. 528. It observed that “[h]e
was a leader of the gang . . . and was a participant and present
at many of the violent activities that occurred here.” Kelly App.
528. A reasonable jurist easily could have imposed the life sen-
tences the District Court did.
4. Schueg
Schueg’s challenges to his concurrent 165-month sen-
tences all relate to the assessment of fines and costs. After stat-
ing simply that Schueg “has the ability to pay a fine,” the Dis-
trict Court ordered that he, together with other defendants, pay
$6,500 in restitution under the Mandatory Victims Restitution
Act (MVRA). Schueg App. 63-64. It also ordered payment of
the special assessment under 18 U.S.C. § 3013(a)(2)(A), and
of certain costs of prosecution, including $13,948.76 for the
compensation of York police officers who testified at trial. Alt-
hough Schueg challenges the MVRA and police compensation
orders on substantive grounds, he also, as a threshold matter,
contests the District Court’s finding of an ability to pay. The
PSR found that Schueg lacked such an ability, and he raised
the issue in his sentencing memorandum.
Under the MVRA, a district court must “specify in the
restitution order the manner in which, and the schedule accord-
ing to which, the restitution is to be paid,” after considering the
defendant’s “financial resources and other assets,” projected
income, and “financial obligations.” 18 U.S.C. § 3664(f)(2).

87
We have interpreted this provision loosely, requiring only that
“the record evidence[] a court’s consideration of the defend-
ant’s financial situation,” though “express findings” need not
be made. United States v. Lessner, 498 F.3d 185, 202 (3d Cir.
2007). Nevertheless, in this case, we cannot find in the record
any consideration of Schueg’s financial condition. There was
testimony regarding a denial of financial aid on a college ap-
plication, and gifts that Schueg gave to his sister’s children.
None of that, however, goes to his ability to pay at the time of
sentencing. While the District Court did specify a payment
schedule, there is no indication where the Court determined
Schueg had the ability to fulfill that schedule—especially
given the PSR’s finding and Schueg’s objection in his sentenc-
ing memorandum. We will, therefore, vacate the District
Court’s judgment of sentence as it relates to the assessment of
restitution, fines, and costs, and remand for consideration of
Schueg’s ability to pay.
5. Atkinson
Atkinson contests the District Court’s application of a
two-level enhancement for obstructing the administration of
justice. To be eligible for that increase, a defendant must (as
relevant here) have “willfully . . . attempted to obstruct or im-
pede[] the administration of justice with respect to the . . . sen-
tencing of the instant offense of conviction.” U.S.S.G. § 3C1.1.
While Atkinson was in prison awaiting sentencing, he alleg-
edly stabbed Carl Hodge, a fellow prisoner, multiple times
while the latter was in the shower. The proximate cause of the
episode, according to Hodge’s testimony at Cruz’s sentencing
hearing, was that Hodge came into possession of a cellphone
Hernandez was using for ongoing illegal activities: bribing
prison guards, selling drugs, and arranging a murder. Hodge
began to share the phone’s contents with the Government. Cruz
and Atkinson became suspicious, leading to the assault.

88
 Atkinson does not dispute Hodge’s testimony. He ar-
gues, rather, that even if he had a motive to harm Hodge be-
cause of suspected cooperation, he could not reasonably have
believed that Hodge would testify against him at sentencing.
See United States v. Galaviz, 687 F.3d 1042, 1043 (8th Cir.
2012). Section 3C1.1 does not demand such a standard. Testi-
mony at sentencing is only one means Hodge could potentially
have disadvantaged Atkinson’s legal position. As the facts
show, Hodge was cooperating with regard to contemporaneous
events, disclosing potentially prejudicial material to the Gov-
ernment. To demand that Atkinson reasonably believed Hodge
would testify against him is unduly limiting and beyond the
text of § 3C1.1. “[T]he administration of justice with respect
to” sentencing encompasses more than witness testimony.
From that perspective, Atkinson’s enhancement must
remain. His “instant offense” was among other things RICO
conspiracy, and Hodge was suspected of (and indeed was) re-
vealing to the Government information related to ongoing con-
certed illicit activities of at least Hernandez, Cruz, and Atkin-
son. That goes directly to the offense of which Atkinson was
convicted and awaiting sentencing. The District Court, then,
did not clearly err in finding a nexus between the attack and
Atkinson’s pending legal proceedings.
B. Collective Challenges
1. Drug Quantity
Rice, Eatmon, and Kelly challenge the District Court’s
drug-quantity attributions pursuant to the Guidelines’ relevant-
conduct provision.43 See U.S.S.G. §§ 1B1.3(a)(1)(B),

43
Villega also seeks to challenge his offense level on this
ground, pointing out that the District Court did not rule on his

89
2D1.1(a). Our review is for clear error. United States v. Perez,
280 F.3d 318, 352 (3d Cir. 2002). “[W]e permit some degree
of estimation in drug conspiracy cases because the government
usually cannot seize and measure all the drugs that flow
through a large drug distribution conspiracy.” United States v.
Diaz, 951 F.3d 148, 159 (3d Cir. 2020) (internal quotation
marks omitted). Nevertheless, information used for sentencing
“must have ‘sufficient indicia of reliability to support its prob-
able accuracy.’” United States v. Miele, 989 F.2d 659, 663 (3d
Cir. 1993) (quoting U.S.S.G. § 6A1.3(a)).
Rice’s PSR recommended a base offense level of 30,
due to a drug-quantity attribution of 280-840 grams of crack.
See U.S.S.G. § 2D1.1(c)(5). The District Court adopted this
recommendation based upon the findings of the jury. Although
the jury’s findings were on a conspiracy-wide basis, the Dis-
trict Court could also, by a preponderance of the evidence, have
incorporated those findings consistent with the relevant-con-
duct standard. See Collado, 975 F.2d at 995.

objections regarding drug quantity, a dangerous-weapon en-
hancement, and relevant conduct for the RICO conspiracy. But
there is good reason for that: Villega’s trial counsel and the
Government agreed, and represented to the District Court at the
sentencing hearing, that the baseline would be an offense level
of 37, which, with a criminal history category of VI, resulted
in a Guidelines range of 360 months to life imprisonment. Vil-
lega’s counsel thereafter raised no objections to the calculation,
and the District Court applied no additional enhancements. The
ultimate sentence was below the agreed-upon range. Contrary
to Villega’s representations on appeal, it is clear that he waived
any challenges to his offense level. See, e.g., United States v.
Ward, 131 F.3d 335, 342-43 (3d Cir. 1997).

90
 As remarked above, Rogers testified that in the conspir-
acy’s early years, he, Atkinson, Eatmon, and Rice all sold crack
they received from Hernandez and Kelly. Rogers agreed that
they were “essentially getting the same quantities or similar
quantities,” App. 3544-45, and he estimated that in this time he
received approximately 1 kilogram of crack from both Hernan-
dez and Kelly. Further, in around 2006-2007, when those sup-
pliers were imprisoned, Rogers said that he, Atkinson, Eatmon,
Sistrunk, and Rice continued to sell drugs together, and that
they mutually facilitated each other’s drug dealing. Rice does
not dispute this testimony, and other evidence indicates his
continued involvement in the conspiracy in the years thereaf-
ter. The District Court did not clearly err in its attribution.
The same goes for Eatmon. He received a base offense
level of 38, on an attribution of 28 kilograms or more of crack.
Rogers testified that for about a year between 2006 and 2007,
he would bring back from New York 500 to 1000 grams of
crack “[e]very couple of days.” App. 3573. He agreed that he
distributed, and that he saw Eatmon and others distribute,
“many kilos of crack” over that time. App. 3575. Further,
Darvin Allen testified that around that same time, for approxi-
mately one to two years, he received from Eatmon about 14
grams of crack a week. Eatmon indicates nothing in the record
to doubt the reliability of this testimony. The attribution of 28
kilograms or more was not clear error.
Finally, Kelly’s challenge fails on a similar basis. His
base offense level, like Eatmon’s, was 38, thanks to an attribu-
tion of 28 kilograms or more of crack. Rogers testified that he
received approximately 1 kilogram of crack from each of Her-
nandez, Cruz, and Kelly in the years after 2002, and, as just
noted, he said that Atkinson, Eatmon, and Rice all received a
similar amount from at least Hernandez and Kelly. There was
also testimony from a high-level South Side supplier, who said

91
that in these years he moved 500 grams to 1 kilogram of crack
a week, including deliveries to Cruz and Hernandez. Further,
Rogers testified that by 2012, Kelly was present when he paid
Hernandez for crack that had been fronted. This indicates
Kelly’s continued active participation in the conspiracy. Fi-
nally, as mentioned above, there was evidence that Kelly con-
tinued to associate with Cruz and Hernandez, and supply crack
even up to the time of the initial indictment in March 2014.
Given this longitudinal evidence of Kelly’s twelve-year partic-
ipation in the highest levels of the conspiracy, the indications
of persistent drug-dealing activity, and the testimony regarding
the amounts involved, we cannot say the District Court clearly
erred in its attribution.
2. Body-Armor Enhancement
During his testimony regarding the early years of the
conspiracy, Rogers said that he saw Hernandez and Kelly
wearing bulletproof vests on multiple occasions at Maple and
Duke Streets. Under the Guidelines, a defendant “convicted of
a drug trafficking crime or a crime of violence” may be eligible
for a two- or a four-level increase to his offense level based on
the use of body armor in the commission of the offense.
U.S.S.G. § 3B1.5(1). The two-level increase applies when “the
offense involved the use of body armor.” U.S.S.G.
§ 3B1.5(2)(A). The four-level one applies if “the defendant
used body armor during the commission of the offense, in prep-
aration for the offense, or in an attempt to avoid apprehension
for the offense.” U.S.S.G. § 3B1.5(2)(B). Kelly received the
latter enhancement; Atkinson and Eatmon the former.
Kelly asserts that Rogers’s testimony does not provide
a sufficient nexus between the wearing of the body armor and
the commission of the offense. The commentary to § 3B1.5 de-

92
fines “use” in part as “active employment in a manner to pro-
tect the person from gunfire.” U.S.S.G. § 3B1.5 cmt. n.1. Kelly
was said to have worn body armor multiple times on Maple and
Duke Streets—the eponymous location of the primary crew of
drug traffickers on the South Side. Further, Rogers’s testimony
was not an offhand remark; it came in the context of a descrip-
tion of the conspiracy’s early years, when Kelly and Hernandez
began supplying crack to Rogers, Atkinson, Eatmon, and Rice.
Kelly, Hernandez, and Cruz would be “standing there on Duke
Street, so you would just buy the drugs from them and then go
sell them on your own.” App. 3546. It was also when Kelly and
Hernandez helped to introduce guns to the South Side, and the
South Side-Parkway rivalry escalated from fistfights to gun-
fights. There is, therefore, a spatial and temporal nexus be-
tween Kelly’s use of the body armor and the commission of the
conspiracy offense. Application of the four-level enhancement
was not clear error.
This same evidence supports the application
§ 3B1.5(2)(A) to Atkinson and Eatmon. We apply to the
Guidelines the ordinary principles of statutory interpretation.
See, e.g., United States v. James, 952 F.3d 429, 433, 439 (3d
Cir. 2020). The provisions here are notably different: while the
four-level enhancement concerns the actions of the defendant,
the two-level one concerns the nature of the offense. The lat-
ter—which encompasses “the offense of conviction and all rel-
evant conduct,” U.S.S.G. § 3B1.5 cmt. n.1 (citing U.S.S.G.
§ 1B1.1 cmt. n.1(I))—need only “involve[]” the use of body
armor. According to Rogers’s testimony, Kelly and Hernan-
dez’s use of the body armor occurred at the time Atkinson and
Eatmon were being supplied by them. Eatmon protests he had
not joined the conspiracy by this point, but he presents no evi-
dence to question the District Court’s judgment.

93
 3. Costs of Prosecution
Seven Defendants—Cruz, Hernandez, Villega, Kelly,
Schueg, Atkinson, Sistrunk, and Eatmon—challenge the Dis-
trict Court’s assessment of a fine to reimburse the City of York
for the overtime wages paid to York police officers who testi-
fied at trial. The Government concedes the issue. We will,
therefore, vacate this aspect of the challengers’ judgments of
sentence.
VII. CONCLUSION
For the foregoing reasons, we will affirm the Defend-
ants’ judgments of conviction, and the judgments of sentence
of Williams and Rice. We will vacate Hernandez’s judgment
of sentence in full, and Schueg’s judgment of sentence as to the
assessment of restitution, fines, and costs. We will remand
those two cases for resentencing proceedings consistent with
this opinion. We will also vacate the judgments of sentence of
Cruz, Villega, Kelly, Atkinson, Sistrunk, and Eatmon as to the
police overtime costs.

94
RESTREPO, Circuit Judge, dissenting.

The District Court issued a sua sponte order closing the
courtroom for jury selection. Appellants were eventually
convicted on various counts related to their involvement in a
local street gang and were sentenced to prison. Among other
issues they raise on appeal, Appellants argue that they are
entitled to a new trial because of the courtroom closure. Due
to the deep roots the right to a public trial has in our history and
its critical importance to the functioning of our criminal justice
system, I would reverse Appellants’ convictions and remand
for a new trial. I respectfully dissent.

I.

Following an extensive investigation conducted by the
United States Bureau of Alcohol, Tobacco, Firearms and
Explosives, a grand jury returned a six-count indictment
against twenty-one defendants. From 2002 to 2014, the
defendants were alleged to have participated in a racketeering
conspiracy, a drug trafficking conspiracy, and drug trafficking
while involved with a York, Pennsylvania street gang. After
nine defendants entered into plea agreements with the
Government, twelve went to trial. Ten of these defendants
(collectively, “Appellants”) now appeal their convictions and
sentences ranging from sixty months to life imprisonment.

On the eve of the trial, the District Court issued an order
closing the courtroom for the entirety of jury selection. In full,
the order states:

AND NOW, on this 18th day of September
2015, IT IS HEREBY ORDERED THAT due
to courtroom capacity limitations, only (1) court

1
 personnel, (2) defendants, (3) trial counsel and
support staff, and (4) prospective jurors shall be
allowed into the courtroom during jury selection.
No other individuals will be present except by
express authorization of the Court.

App. 10 (bold in original). Neither the Government nor the
defendants requested this order, and the District Court did not
seek their input. The Court closed the courtroom to the public
without determining whether it was necessary or considering
any alternatives. None of the defendants objected to the order,
and voir dire then took place for two days.

II.

We must now decide whether to correct an erroneous
courtroom closure despite Appellants’ failure to object. As a
preliminary matter, it is imperative to understand the contours
of the constitutional right in question.

The Sixth Amendment provides that “the accused shall
enjoy the right to a speedy and public trial”—and the Supreme
Court has long recognized the importance of the public trial
right for the accused and the broader community. See, e.g., In
re Oliver, 333 U.S. 257 (1948). “[T]he Sixth Amendment right
to a public trial extends to the voir dire of prospective jurors.”
Presley v. Georgia, 558 U.S. 209, 213 (2010) (per curiam); see
also Press-Enterprise Co. v. Superior Court, 464 U.S. 501, 508
(1984) (noting that “the accused’s right [to a fair trial] is
difficult to separate from the right of everyone in the
community to attend the voir dire” under the First
Amendment). As a part of the public trial right, criminal
defendants and the public at large are entitled to open
proceedings.

2
 The public trial guarantee is deeply rooted in our
common law heritage. In England, early court proceedings
required public access to “moots,” which later evolved into
juries, consisting of “the freemen of the community.” See
Press-Enterprise, 464 U.S. at 505. In the eleventh century, the
jury began to transform into a small group of individuals that
represented the community, but “the public character of the
proceedings, including jury selection, remained unchanged.”
Id. at 506. As early as the sixteenth century, jurors in England
were selected “openlie in the presence of the Judges, the
Justices, the enquest, the prisoner, and so many as will or can
come so neare as to heare it.” Id. at 507 (emphasis in original)
(quoting Thomas Smith, De Republica Anglorum 96 (1565)
(Alston ed. 1906)).

The presumption of public jury selection “carried over
into proceedings in colonial America.” Id. at 508 (discussing
accounts on the need for bystanders at trials following the
Boston Massacre). Many of the thirteen colonies enacted
statutes requiring jury selection to occur in open court. See id.
(“Public jury selection . . . was the common practice in
America when the Constitution was adopted.”). For instance,
a 1773 statute in North Carolina required that court clerks,

write the Names of all Petit Jurors appearing, on
Scrolls or Pieces of Paper, which shall be put into
a Box; and on every Issue in every Suit where it
is not otherwise agreed by Consent, a Child
under Ten Years old, in open Court, shall draw
out of the said Box Twelve of the said Scrolls or
Pieces of Paper.

James Davis, Complete Revisal of All the Acts of Assembly, of
the Province of North-Carolina, Now in Force & Use 549

3
(1773) (emphasis added). Delaware employed a similar
system in which prospective jurors’ names were placed in a
box until “some indifferent person, by the direction of the
court, may and shall, in open court, draw out twelve of the said
pieces of parchment or paper.” 2 Laws of the State of Delaware
1073 (Samuel & John Adams eds. 1797) (emphasis added).
These are just two examples, as open voir dire proceedings
were the practice at the time of our Nation’s founding.

The Sixth Amendment enshrined the presumption of
public access in the Constitution. The Founding Fathers
believed that public court proceedings provided safeguards
integral to the nascent republic. At the Constitutional
Convention, broad agreement existed regarding the jury trial’s
importance as “a valuable safeguard to liberty . . . [or] the very
palladium of free government.” The Federalist No. 83, at 461
(Alexander Hamilton) (P.F. Collier ed., 1901). And jury
selection was viewed as a “double security” against corruption
that would require a person to “corrupt both the court and the
jury.” Id. at 463.

Enunciating “revolution principles” under the
pseudonym “Novanglus,” John Adams struck similar themes
when he explained that “draw[ing] [jurors] by chance out of a
box in open town meeting” best “secured against a possibility
of corruption of any kind . . . having seen with their own eyes,
that nothing unfair ever did or could take place.” John Adams,
Novanglus; or, A History of the Dispute with America from Its
Origin, in 1754, to the Present Time, The Revolutionary
Writings of John Adams 152, 199 (C. Bradley Thompson, ed.,
2000) (emphasis added). These sentiments were explicitly
incorporated into the Constitution in the language of the Sixth
Amendment.

4
 It is thus no surprise that the Supreme Court classifies
courtroom closures “as a structural error” that generally
“entitl[es] the defendant to automatic reversal.”1 Weaver v.
Massachusetts, 137 S. Ct. 1899, 1905 (2017) (plurality
opinion). Courts usually reverse criminal convictions tainted
by a structural error because they affect “the framework within
which the trial proceeds,” thus “infect[ing] the entire trial
process” and undermining the ultimate determination of “guilt
or innocence.” Neder v. United States, 527 U.S. 1, 8–9 (1999)
(citations and internal quotation marks omitted). An open
courtroom during jury selection is fundamental to protecting
defendants’ right to a jury free from prejudice and ensuring

1
There are limited instances in which closing a
courtroom is not structural error. A judge may order a
closure based on findings that specifically identify “higher
values” that must be preserved. Waller v. Georgia, 467 U.S.
39, 45 (1984) (quoting Press-Enterprise, 464 U.S. at 510);
see also Weaver, 137 S. Ct. at 1909 (“[A] judge may deprive
a defendant of his right to an open courtroom by making
proper factual findings in support of the decision to do so.”).
Trial courts are required to “consider alternatives to closure
even when they are not offered by the parties.” Presley, 558
U.S. at 214.

The District Court did not consider alternative options
or make any factual findings in support of its order. The
Government points to comments the District Court made on
the number of people in the courtroom. However, these
comments do not support the proposition that the District
Court made the required findings because they came days
after the order and are not linked in any discernible way to the
closure.

5
public confidence in the administration of justice. See Press-
Enterprise, 464 U.S. at 508; United States v. Negrón-Sostre,
790 F.3d 295, 301 (1st Cir. 2015). Accordingly, it was a
structural error to close the courtroom during voir dire.

III.

There are instances in which a structural error does not
automatically lead to a reversal. In Weaver, the Supreme Court
recently examined an erroneous courtroom closure on
collateral review. Due to space limitations, “an officer of the
court excluded from the courtroom any member of the public
who was not a potential juror.” Weaver, 137 S. Ct. at 1906.
Citing finality concerns, the plurality concluded that the
petitioner did not demonstrate prejudice as required for a new
trial under Strickland v. Washington, 466 U.S. 668 (1984). See
Weaver, 137 S. Ct. at 1912–14. Although the Weaver plurality
cautioned courts not to assume that public trial violations
always require reversal in a collateral proceeding, it did not
address the appropriate remedy when the error is raised for the
first time on direct review.

Here, Appellants did not object to the District Court’s
closure order or otherwise preserve their claim during the trial.
We thus review the order for plain error. Fed. R. Crim. P.
52(b). Appellants must satisfy four prongs under plain error
review. See United States v. Olano, 507 U.S. 725, 736 (1993).
They must show (1) that there was an error, (2) that was “clear
or obvious,” and (3) it must have impacted their “substantial
rights.” Puckett v. United States, 556 U.S. 129, 135 (2009).
The third prong generally “means that the error must have been
prejudicial,” meaning “[i]t must have affected the outcome of
the district court proceedings.” Olano, 507 U.S. at 734.

6
Fourth, reviewing courts have discretion to remedy a forfeited
error if it “seriously affect[ed] the fairness, integrity or public
reputation of judicial proceedings.” Id. at 736 (citation and
internal quotation marks omitted). Only the third and fourth
prongs are relevant for our purposes because the parties agree
that the closure was a clear error. Below, I will consider these
prongs in turn.

A.

Olano’s substantial rights prong typically requires a
showing of prejudice. Puckett, 556 U.S. 129. “To satisfy this
. . . condition, the defendant ordinarily must show a reasonable
probability that, but for the error, the outcome of the
proceeding would have been different.” Rosales-Mireles v.
United States, 138 S. Ct. 1897, 1904–05 (2018) (citation and
internal quotation marks omitted). But “[t]here may be a
special category of forfeited errors that can be corrected
regardless of their effect on the outcome.” Olano, 507 U.S. at
735. The Majority declines to address whether an erroneous
courtroom closure fits this “special category” under the third
Olano prong. Majority Op. at 17 (noting that it need not decide
because it declines to exercise discretion under the fourth
prong). I disagree and would hold that the specific structural
error at issue here fits the special category of errors that must
be corrected even without a particularized showing of
prejudice and thus satisfies Olano’s third prong.

The Supreme Court has made clear that structural errors
generally result in the reversal of a conviction because they
“are so intrinsically harmful as to require automatic reversal
(i.e., ‘affect substantial rights’) without regard to their effect
on the outcome.” Neder, 527 U.S. at 7. Requiring defendants

7
to make a specific showing of prejudice when claiming a
structural error on direct review would force them to engage in
a “speculative inquiry into what might have occurred in an
alternate universe.” United States v. Gonzalez-Lopez, 548 U.S.
140, 148–50 (2006) (describing why it is “unnecessary to
conduct an ineffectiveness or prejudice inquiry” to establish a
violation to the “right to counsel of choice”).

The District Court’s closure of the courtroom during
voir dire is the prototypical constitutional error that is
impossible to measure. “Jury selection is the primary means
by which a court may enforce a defendant’s right to be tried by
a jury free from ethnic, racial, or political prejudice . . . , or
predisposition about the defendant’s culpability.” Gomez v.
United States, 490 U.S. 858, 873 (1989). Public jury selection
proceedings impact the way in which potential jurors respond
to questions about their past experiences and the types of
questions attorneys ask them. See Negrón-Sostre, 790 F.3d at
305–06.

The difficulty in determining the level of prejudice is
precisely why structural errors are presumed to affect
defendants’ substantial rights. See Neder, 527 U.S. at 7.
Contrary to the Majority, I do not view the conclusion that the
District Court’s courtroom closure affected Appellants’
substantial rights as a “doctrinal leap.” See Majority Op. at 17.
It would be illogical to classify an error as structural because it
affects substantial rights but then conclude that it did not affect
defendants’ substantial rights for purposes of Olano’s third
prong. Given the difficulty of measuring prejudice arising
from a public trial violation and the importance of jury
selection in protecting criminal defendants, this Court should
presume prejudice and hold that Appellants have satisfied the
substantial rights prong.

8
 B.

The District Court’s order also undermines the fairness,
integrity, and public reputation of the trial proceedings, thus
satisfying Olano’s fourth prong. As explained above, open
voir dire is key to ensure that unprejudiced jurors are
ultimately selected to serve on juries. It also serves as a check
on judicial abuse against defendants caught up in the criminal
justice system. See United States v. Lnu, 575 F.3d 298, 305
(3d Cir. 2009) (stating that the public trial right “has always
been recognized as a safeguard against any attempt to employ
our courts as instruments of persecution”) (citation and internal
quotation marks omitted). Even in cases where there are no
further constitutional violations, open jury selection maintains
the public’s confidence in the system by enhancing “the
appearance of fairness.”2 Press-Enterprise, 464 U.S. at 508;
see also Waller, 467 U.S. at 46 (stating that public trials ensure
that the “judge and prosecutor carry out their duties
responsibly” and “discourages perjury”).3

2
These fairness concerns are particularly relevant in
light of the District Court’s handling of the Batson challenge
in chambers. Although I agree with the Majority that the
resolution of the challenge in camera was harmless, the
District Court’s conduct is concerning because it represents
another instance in which the Court limited access to jury
selection proceedings.
3
Concerns related to public confidence in the
proceedings are especially relevant here given the local media
coverage into the case. See, e.g., Keith Schweigert, York
member of Southside gang to serve 21 years on drug,
racketeering charges, Fox 43 (December 21, 2018, 11:24

9
 The pivotal role that public proceedings play in our
judicial system is precisely why reviewing courts find it
particularly problematic when trial judges themselves limit
access to courtrooms. See Weaver, 137 S. Ct. at 1913
(emphasizing that the “closure decision . . . was made by court
officers rather than the judge”). It is also why trial judges are
responsible for considering alternatives to closure even if none
are raised by the parties. Presley, 558 U.S. at 214–15 (noting
that trial courts must consider alternatives given jury
selection’s importance “to the adversaries [and] to the criminal
justice system”) (citation and internal quotation marks
omitted). As the reviewing court, it is imperative that we
correct the District Court’s structural error because it
undermines the integrity and public reputation of criminal
proceedings that resulted in Appellants’ convictions.

Instead, the Majority conducts a cost-benefit analysis to
justify leaving the public trial violation uncorrected. Majority
Op. at 29 (declining remedial action because “the remedy is to
be assessed relative to the costs of the error”). This approach
is foreign and detrimental to our structural error jurisprudence.

The Majority first minimizes the impact of the error by
pointing out that there is no evidence anyone sought to access
to the courtroom, that there is no indication of wrongdoing by
the District Court or the Government, and that transcripts of

AM), https://fox43.com/2018/12/21/york-member-of-
southside-gang-to-serve-21-years-on-drug-racketeering-
charges/; Christopher Dornblaser, Life in prison for York City
Southside gang leader, York Dispatch (October 3, 2017, 8:03
PM),
https://www.yorkdispatch.com/story/news/2017/10/03/life-
prison-york-city-southside-gang-leader/729170001/.

10
voir dire were made available. Majority Op. at 26–28. The
availability of transcripts does little to mitigate the error
because “no transcript can recapture the atmosphere of the voir
dire, which may persist throughout the trial.” Gomez, 490 U.S.
at 874–75; see also United States v. Antar, 38 F.3d 1348, 1360
n.13 (3d Cir. 1994) (explaining that “the translation of a live
proceeding to a cold transcript” misses “some information,
concerning demeanor, non-verbal responses, and the like”).

The other two factors the Majority mentions miss the
point of structural errors like public trial violations. Much of
the Majority’s analysis relies on cases that consider errors
reviewed for harmlessness. See Majority Op. at 23–25. At one
point, the Majority even posits that “apart from cases of actual
innocence, an altered outcome does not in itself necessitate
correction of the error.” Majority Op. at 25. The Majority
overlooks the critical fact that we do not review criminal trials
with a structural error for harmlessness because such trials
“cannot reliably serve its function as a vehicle for
determination of guilt or innocence, and no criminal
punishment may be regarded as fundamentally fair.” Arizona
v. Fulminante, 499 U.S. 279, 310 (1991) (citation and internal
quotation marks omitted). Because public trial violations
corrupt the very mechanism used to determine guilt or
innocence, we cannot measure the true costs of leaving the
District Court’s error uncorrected.

The Majority next focuses on the high costs of remedial
action to correct the error. Correcting the public trial violation
would require reversal of Appellants’ convictions, which
resulted from two-month long proceedings completed five
years ago, and remand for a new trial. The costs to remedy the
District Court’s error are indeed considerable. I disagree,
however, with the central role the Majority affords these costs

11
in its plain error analysis. The District Court committed a
grave constitutional violation by simultaneously violating
twelve defendants’ right to a public trial for the entirety of jury
selection. The nature of the error, not the cost of correcting it,
must be the lodestar of our consideration of a structural error
on plain-error review. The District Court “undermine[d] the
structural integrity of the criminal tribunal itself” in a way that
“is not amenable to harmless-error review”—and the Majority
allows this to stand. Vasquez v. Hillery, 474 U.S. 254, 263–64
(1986). It is perverse to weigh the costs of judicial efficiency
against Appellants’ constitutional rights when the District
Court undeniably committed structural error.

For the reasons stated above, I respectfully dissent. A
balancing test or a cost-benefit analysis is an improper and
unjust method for determining whether to protect certain
fundamental constitutional rights. The public trial right is one
of these fundamental rights. It has deep roots in our Nation’s
history and is essential to the functioning of our criminal
justice system. I would therefore reverse Appellants’
convictions and remand for a new trial.

12